Freeze-dried snake antivenoms formulated with sorbitol, sucrose or mannitol: Comparison of their...

transcript

ilable at ScienceDirect

Toxicon 90 (2014) 56e63

Contents lists ava

Toxicon

journal homepage: www.elsevier .com/locate/ toxicon

Freeze-dried snake antivenoms formulated with sorbitol,sucrose or mannitol: Comparison of their stability in anaccelerated test

María Herrera a, *, Virgilio Tattini Jr. b, Ronaldo N.M. Pitombo b,Jos�e María Guti�errez a, Camila Borgognoni b, Jos�e Vega-Baudrit c,Federico Solera c, Maykel Cerdas a, �Alvaro Segura a, Mauren Villalta a,Mari�angela Vargas a, Guillermo Le�on a

a Instituto Clodomiro Picado, Facultad de Microbiología, Universidad de Costa Rica, San Jos�e, Costa Ricab Department of Biochemical and Pharmaceutical Technology, School of Pharmaceutical Sciences, University of Sao Paulo, Av. Prof. LineuPrestes, 580, Bloco 16, CEP05508-900, Sao Paulo, SP, Brazilc Laboratorio Nacional de Nanotecnología, LANOTEC-CENAT, San Jos�e, Costa Rica

a r t i c l e i n f o

Article history:Received 31 March 2014Received in revised form 22 July 2014Accepted 24 July 2014Available online 1 August 2014

Keywords:AntivenomFreeze-dryingSorbitolMannitolSucroseStability

* Corresponding author. Tel.: þ506 2511 7878; faE-mail addresses: maria.herrera_v@ucr.ac.cr,

com (M. Herrera).

a b s t r a c t

Freeze-drying is used to improve the long term stability of pharmaceutical proteins. Sugarsand polyols have been successfully used in the stabilization of proteins. However, their usein the development of freeze-dried antivenoms has not been documented. In this work,whole IgG snake antivenom, purified from equine plasma, was formulated with differentconcentrations of sorbitol, sucrose or mannitol. The glass transition temperatures of frozenformulations, determined by Differential Scanning Calorimetry (DSC), ranged between�13.5 �C and �41 �C. In order to evaluate the effectiveness of the different stabilizers, thefreeze-dried samples were subjected to an accelerated stability test at 40 ± 2 �C and75 ± 5% relative humidity. After six months of storage at 40 �C, all the formulations pre-sented the same residual humidity, but significant differences were observed in turbidity,reconstitution time and electrophoretic pattern. Moreover, all formulations, except anti-venoms freeze-dried with mannitol, exhibited the same potency for the neutralization oflethal effect of Bothrops asper venom. The 5% (w:v) sucrose formulation exhibited the beststability among the samples tested, while mannitol and sorbitol formulations turnedbrown. These results suggest that sucrose is a better stabilizer than mannitol and sorbitolin the formulation of freeze-dried antivenoms under the studied conditions.

1. Introduction

Freeze-drying or lyophilization is the most commonlyused method for preparing solid proteins which are phys-ically or chemically unstable in aqueous solution. However,

x: þ506 2292 0485.mariaherrv@hotmail.

most proteins are sensitive to lyophilization due to thestress of freezing and drying that can cause irreversibledamage to the protein structure and biological activity(Heller et al., 1999; Sarciaux et al., 1999). The effectivenessof this technology in the stabilization of biopharmaceuticalproducts is the result of the combination of optimizing theformulation and controlling the process. Formulationoptimization is focused on the use of disaccharides as sta-bilizers, together with bulking agents, such asmannitol andglycine (Imamura et al., 2003; Sharma and Kalonia, 2004).

M. Herrera et al. / Toxicon 90 (2014) 56e63 57

On the other hand, the optimization of the process involvescontrolling the freezing and drying stages for each formu-lation developed.

Immunoglobulins are a group of proteins relevant aspharmaceuticals and as diagnostic agents. These proteinsare prone to forming aggregates and undergo other phys-ical and chemical modifications during manufacturing andlong term storage, which may cause loss of activity (Mauryet al., 2005). The immunoglobulins G (IgGs) constitute theactive principle of antivenoms. Snake antivenoms areconsidered the only scientifically proven therapy againstsnakebite envenomation (Bon, 1996), and they are pro-duced from the plasma of animals immunized with avenom or a mixture of venoms.

Snakebite envenomation is an important neglectedtropical disease in many regions of the world, particularlysub-Saharan Africa, Asia, Latin America and Papua-NewGuinea (Guti�errez et al., 2011). The global crisis of anti-venom supply and the need to distribute antivenoms inremote areas of developing countries, where an adequatecold chain cannot be guaranteed, underscore the impor-tance of freeze-dried formulations for snake antivenoms,ensuring their stabilization during transport and storage.

Despite the need to produce more stable and easy todistribute antivenoms, this issue has received little atten-tion by antivenoms manufacturers, and although manycommercial formulations are freeze-dried, there is a verylimited body of published literature on freeze-drying ofsnake antivenoms. To the best of our knowledge, thethermal properties of antivenoms and their stability afterfreeze-drying and during storage in the solid state have notbeen reported.

The thermal characterization of antivenom formula-tions, and particularly the determination of the glasstransition temperature of the maximally freeze-concentrated solution (Tg'), are critical parameters in thedevelopment of the freeze-drying cycle (Wang, 2000). Tg'defines the maximally allowable temperature for the pri-mary drying since, if the product temperature exceeds thiscritical temperature, amorphous collapse could occur(Kasper and Friess, 2011). The stability and activity offreeze-dried antibodies largely depend on the processingconditions and on the use of an adequate stabilizer at theoptimum concentration (Sarciaux et al., 1999).

In this work, we performed a thermal characterizationof snake antivenoms formulations, assessed the effect offreeze-drying on equine antibodies, and evaluated theeffectiveness of sorbitol, sucrose andmannitol as stabilizersin antivenom samples subjected to an accelerated stabilitytest during six months.

2. Materials and methods

2.1. Snake venom

Pools of venoms from adult specimens of the snakesBothrops asper, Crotalus simus and Lachesis stenophrys,maintained in captivity at the Serpentarium of InstitutoClodomiro Picado, were obtained by mechanical stimula-tion of venom glands, stabilized by freeze-drying and keptat �20 �C until use. For neutralization studies, only the

venom of B. asper was used, since this species is the mostimportant venomous snake in Central America.

2.2. Antivenoms production and formulation with stabilizers

Plasma from horses immunized with a mixture of thevenoms of B. asper, C. simus and L. stenophrys was used as astarting material. The antivenom immunoglobulins werepurified by precipitation with 5% caprylic acid, followed byvigorous agitation for 1 h. Then the immunoglobulins weremicro-filtered through an 8 mm retentive paper (WhatmanN� 2, Kent, UK), dialyzed against distilled water, andformulated with deionized water at a total protein con-centration of 8 g/dL and a pH of 7.0 (Rojas et al., 1994).Additionally, antivenoms were formulated with either0.05M, 0.5M,1M or 2M sorbitol (Sigma S-7547), 2% or 10%mannitol (Merck-5982), 2% or 5% sucrose (Sigma S-5016) or0.9% NaCl (Sigma S-1679). Antivenom without excipientwas used as a control.

2.3. Differential Scanning Calorimetry (DSC)

The glass transition temperature (Tg') was determinedusing a differential scanning calorimeter, model Q200 (TAInstruments, USA). Twenty mg of each antivenom formu-lation were placed in an aluminum pan that was hermeti-cally sealed and frozen at �50 �C at scan rates of 5e10 �C/min, followed by an isotherm of 3 min, and heated at 25 �Cat scan rates of 2.5e5.0 �C/min. All the Tg' were reported asthe midpoint of the transition.

2.4. Freeze drying microscopy (FDM)

Collapse temperatures were measured using a freezing-drying cryo-stage FDCS 196 (Linkam Scientific Instruments,UK) equipped with a liquid nitrogen cooling system LNP94/2 (Linkam Scientific Instruments, UK), a programmabletemperature controller, and a vacuum pump EdwardsE2M1.5 (Linkam Scientific Instruments, UK). Samples wereplaced on a 16 mm quartz cover slip and were frozen to�50 �C at 10 �C/min. Each sample was heated under vac-uum (about 1 Pa) at 3 �C/min up to 0 �C. Direct observationof microscopic collapse was done by using a Nikon EclipseE600 (Nikon, Japan) polarized microscope with acondenser extension lens.

2.5. Freeze-drying of antivenoms

Ten milliliter vials were filled with 5 mL of eachformulation and loaded on a freeze-dryer Benchmark 1100(Virtis, USA). The samples were frozen at �40 �C andannealed at �10 �C for 4 h. The primary drying was con-ducted at�20 �C for 64 h, and the secondary drying at 30 �Cfor 4 h and 200 mTorr.

2.6. Residual moisture

Residual moisture was measured by the Karl Fishertitration method using a TitroLine KF apparatus (Schott,Germany). At least 50 mg of pulverized antivenom cakewere dispersed in a methanol bath (Merck, LiChrosolv®

Table 1Glass Transition Temperature (Tg') of frozen snake anti-venom formulations.

Formulation Tg' (�C)a

Control �13.5 ± 0.1Sorbitol 0.05 M �23.6 ± 0.0Sorbitol 0.5 M �23.0 ± 1.0Sorbitol 1.0 M �41.3 ± 0.6Sorbitol 2.0 M �41.4 ± 1.3Mannitol 2% �19.0 ± 3.0Mannitol 10% �32.1 ± 0.2Sucrose 2% �22.9 ± 0.4Sucrose 5% �22.0 ± 4.0NaCl 0.9% �22.6 ± 1.4

a Results are presented as mean ± S.D. (n ¼ 3).

M. Herrera et al. / Toxicon 90 (2014) 56e6358

106018) and titrated with Karl Fischer reagent (Fisher Sci-entific AL2000-1) until the end point was reached, asdetermined by the KF processor.

2.7. Reconstitution time

Freeze-dried antivenom samples were reconstitutedwith 10 mL of 0.85% saline solution (sterile and non-pyrogenic), and the dissolution was observed visually asthe vials were gently agitated by hand. The time required toachieve complete dissolution was recorded.

2.8. Turbidity analysis

Turbidity of the preparations was quantified using aturbidimeter model 2020 (La Motte, USA) that was cali-brated with standards (HACH Company, USA) prior toanalysis. Turbidity was expressed as nephelometricturbidity units (NTU).

2.9. Neutralization of lethality assay

The neutralization of lethality activity was performedonly for B. asper venom. Mixtures containing a constantamount of this venom, corresponding to fourMedian LethalDoses (LD50), i.e. four times theamountof venomrequired tokill 50%of the animals, andvariousdilutions of antivenom in0.12MNaCl, 0.04M phosphate, pH 7.2 (PBS), were preparedand incubated for 30min at 37 �C. Controls included venomincubatedwith PBS instead of antivenom. Aliquots of 0.5mLof the mixtures were injected, by the intraperitoneal route,to groups offive CD-1mice (16e18 g). Deathswere recordedduring 48 h, and neutralizing activity, expressed as MedianEffectiveDose (ED50)was estimated byProbits (Solano et al.,2010). The experimental protocols involving animals wereapproved by the Institutional Committee for the Care andUse of LaboratoryAnimals (CICUA) of theUniversity of CostaRica (Project 82-08) and meet the International GuidingPrinciples for Biomedical Research Involving Animals(CIOMS, 1986).

2.10. Stability study

The thermal stability of the antivenoms formulatedwithdifferent excipients was assessed by incubating the sam-ples at 40 �C ± 2 �C and at a relative humidity of 75% ± 5during six months. Samples of each formulation wereanalyzed at the beginning and the end of the study forappearance of the cake, residual moisture content, recon-stitution time, turbidity, electrophoretic pattern andneutralization of the lethal activity of B. asper venom.

2.11. Electrophoretic analysis

Twenty mg of total protein of each formulation wereseparated by SDS-polyacrylamide gel electrophoresis (SDS-PAGE) under non-reducing conditions, using an acrylamideconcentration of 7.5%. Gels were stained with CoomassieBrilliant Blue R-250. The starting voltagewas 180 V and gelswere destained with a mixture of methanol, ethanol andacetic acid.

2.12. Total protein determination

Total protein concentration was determined by amodification of the Biuret test (Parvin et al., 1965) in which50 mL of samples (or protein standards) were mixed with2.5 mL of Biuret reagent and incubated at room tempera-ture during 30 min. Absorbances at 540 nmwere recordedand protein concentration was calculated using the equa-tion of the curve obtained by plotting the absorbance of thestandards as a function of their protein concentration.

2.13. Statistical analysis

Results were expressed as mean ± SD from triplicates.The statistical analysis was performed using the softwareGraphPad InStat™ (Goteborg, Sweden). The significance ofthe differences between the mean values of experimentalgroups was determined by ANOVA followed by theTukeyeKramer test. Differences were considered statisti-cally significant at values of P < 0.05. Comparison of theED50 values was performed using the 95% confidencelimits. Values were considered significantly different whentheir 95% confidence limits did not overlap.

3. Results and discussion

3.1. Thermal analysis of frozen formulations

DSC and FDM have been successfully used to determineglass transition and collapse temperatures of therapeuticproteins (Meister and Gieseler, 2009). In this work, theglass transition temperature of themaximally concentratedmatrix (Tg') was determined by DSC, as shown in Table 1.

TheDSC curve for all frozen formulations showed a slightendothermic shift in heat capacity, associated with glasstransition, as well as an endothermic curve close to 0 �Ccorresponding to the melting of ice (data not shown).Moreover, the curve for 10% mannitol formulation alsoshowed an exothermic event in the range of �26 �C to�19 �C, suggesting the further and probable completion ofmannitol crystallization during warming. This completionofmannitol crystallization is known as themetastable state.

The glass transition temperature for control antivenom,calculated at the midpoint of transition, was �13.5 �C, avalue close to the Tg' for other pure proteins like myoglobin

(�10.5 �C), a-casein (�12.5) and ovalbumin (�11 �C), asreported by Wang (2000). Moreover, the glass transitiontemperature of the other formulations was lower than theTg' of control antivenom, and in the formulations where theexcipient was the most abundant component, Tg' was closeto the value of the pure excipient (Wang, 2000). In anti-venoms formulated with 1 M sorbitol and 2 M sorbitol, theTg' value was very low; thus, these formulations wereanalyzed by freeze-drying microscopy to determine thecollapse temperature and to confirm the critical tempera-ture of the formulations. On the other hand, a thermaltreatment or annealing step was introduced on the freeze-drying cycle to allow complete crystallization of mannitoland removal of the metastable state.

3.2. FDM of formulations with sorbitol

The collapse temperature of the antivenoms formulatedwith 1 M and 2 M sorbitol was determined by FDM (Fig. 1).FDM allows the direct observation of separation betweenfrozen and dried regions of the sample, due to viscous flowof the amorphous phase during heating. This method runsunder different experimental conditions to those used inDSC and, therefore, they are considered complementarymethodologies in studying thermal properties of proteinformulations.

Usually, collapse temperatures are in the range of2e5 �C above glass transition temperatures (Pikal and Shah,1990); however, in this study the collapse temperature didnot show a trend, being �31 �C and �41 �C for antivenomsformulated with 1 M and 2 M sorbitol, respectively. Sincecollapse temperatures are very low in both samples, theprimary drying should be conducted at very low temper-ature and, consequently, the lyophilization process wouldbe very prolonged. Therefore these antivenom formula-tions were not considered suitable for freeze-drying.

3.3. Characterization of freeze-dried formulations

Only the antivenoms formulated with 0.05 M sorbitol,0.5 M sorbitol, 2% mannitol, 10% mannitol, 2% sucrose, 5%sucrose and 0.9% NaCl were submitted to freeze-drying andfurther analysis. The visual examination of the freeze-dried

Fig. 1. Freeze-drying microscopy profile of the antivenom formulated with 1.0 M sothe collapse of the freeze-drying front (B).

formulations did not show any sign of collapse or shrinkageof the cakes, with the exception of the antivenom formu-lated with 0.9% NaCl. The macroscopic collapse exhibitedby this sample was considered unacceptable and thisformulation was discarded from further studies. In Table 2,the residual humidity, reconstitution time, turbidity andpotency of the freeze-dried antivenomswere compared. Allthe antivenoms with sorbitol, sucrose or mannitol, exceptthe formulation with 10% mannitol, displayed a lowerturbidity than control antivenoms (P < 0.05). The effect ofpolyols for decreasing turbidity of liquid snake antivenomshas already been reported (Segura et al., 2009), as well asthe decrease of aggregates in freeze-dried and spray-driedproteins stabilized with sucrose and sorbitol (Maury et al.,2005; Wang et al., 2009).

The reconstitution time of freeze-dried antivenoms isrelevant at the clinical setting for two main reasons: (a)because the elapsed time between the snakebite and thebeginningof thetherapyhasan impactontheevolutionof theenvenomation (Oteroetal., 2002)and(b)becausedifficulty inreconstitution of antivenomsmay reflect denaturation of theantibodies, implying loss of stability and neutralizing activity(Theakston et al., 2003). Previous studies have reported pro-longed reconstitution times of 30 and 90 min for somelyophilized antivenoms (Hill et al., 2001; Quan et al., 2010),which might affect the efficacy of treatment, while othersrecommend different strategies of reconstitution to improvethe dissolution time (Gerring et al., 2013). In thiswork, all theformulations showed reconstitution times lower than 5 minin accordance with times reported for other freeze-driedproteins (Searles et al., 2001; Schersch et al., 2010), corrobo-rating that it is possible to produce antivenoms of easyreconstitution without loss of activity.

The use of sorbitol, mannitol or sucrose in the freeze-drying of antivenoms did not affect the residual moisturecontent of the samples, since it was lower than 5% for allformulations, with the exception of the control. To the bestof our knowledge, the appropriate residual moisture forsnake antivenoms has not been reported. It has been sug-gested that lower moisture content leads to more stableprotein products (Wang, 2000).

The effect of freeze-drying on the neutralizing potencyof antivenom antibodies was assessed by a standard in vivo

rbitol showing primary drying under vacuum with the drying front (A), and

Table 2Physicochemical and biological characterization of snake antivenom formulations after freeze-drying.

Formulation Reconstitution time (min)a Potency (mg/mL)b Residual moisture (%)a Turbidity (NTU)a,c

Sorbitol 0.05 M 3.0 ± 0.4 2.14 (1.52e3.01) 3.9 ± 0.4 18.3 ± 0.6Sorbitol 0.5 M 2.9 ± 0.3 1.38 (1.00e1.91) 3.6 ± 0.2 20.0 ± 0.0Controld 2.9 ± 0.2 1.87 (1.43e2.45) 5.7 ± 1.0 22.0 ± 0.0

Mannitol 2% 0.7 ± 0.3 3.12 (2.33e4.17) 3.1 ± 0.5 22.0 ± 0.0Mannitol 10% 1.7 ± 0.4 3.39 (2.54e4.53) 2.9 ± 0.3 24.0 ± 0.0Sucrose 2% 0.7 ± 0.3 2.26 (1.80e2.83) 3.4 ± 0.2 21.0 ± 0.0Sucrose 5% 1.3 ± 0.4 2.26 (1.69e3.02) 3.5 ± 0.6 21.3 ± 0.6Controle 2.6 ± 0.3 2.26 (1.69e3.02) 4.0 ± 0.7 24.0 ± 0.0

a Results are presented as mean ± S.D. (n ¼ 3).b Expressed as mg venom neutralized per mL antivenom; 95% confidence intervals are depicted in parenthesis.c NTU: Nephelometric Turbidity Units.d Control for antivenoms freeze-dried with sorbitol.e Control for antivenoms freeze-dried with mannitol and sucrose.

test in mice using the intraperitoneal route of injection. Forall formulations, the neutralizing potency against B. aspervenom was preserved after freeze-drying. This result sug-gests that antibodies are not affected by the process or bythe presence of excipients; however, since the potencyassay has a normal variation of 30% (Solano et al., 2010), it ispossible that small losses in the potency of the antivenomcannot be detected with this test. Because all formulationsmaintained their physicochemical and biological propertiesafter freeze-drying, these samples were subjected to astability study under extreme conditions of temperatureand relative humidity during six months.

3.4. Stability study of the freeze-dried formulations

During the development of solid protein preparations,the choice of the final formulation is based on the resultsobtained after the freeze-drying process, but especially onthe results generated from stability studies. Acceleratedstability studies involve elevated conditions of temperatureand humidity, established from the normal conditions ofproduct storage and the climatic zone in which the study isconducted. An important issue about accelerated stabilitystudies is whether the data obtained at high temperaturescan be extrapolated to those obtained at real time condi-tions. In this sense some authors have found that, in pro-teins where degradation pathways can be described

Table 3Physicochemical and biological characterization of freeze-dried antivenom form

Formulation Reconstitution time (min)a Potency (m

Sorbitol 0.05 M 10.3 ± 0.3 1.50 (1.20eSorbitol 0.5 M 11.8 ± 0.5 1.08 (0.79eControle 12.6 ± 0.4 1.28 (0.90e

Mannitol 2% 5.4 ± 0.3 1.77 (1.41eMannitol 10% 7.1 ± 0.4 1.92 (1.48eSucrose 2% 8.8 ± 0.3 1.63 (1.26eSucrose 5% 5.9 ± 0.2 2.25 (1.68eControlf 13.4 ± 0.3 2.88 (2.15e

a Results are presented as mean ± S.D. (n ¼ 3).b Expressed as mg venom neutralized per mL antivenom; 95% confidence intec NTU: Nephelometric Turbidity Units.d Not determined.e Control for antivenoms freeze-dried with sorbitol.f Control for antivenoms freeze-dried with mannitol and sucrose.

separately and the rate-limiting degradation step does notchange within a certain temperature range, the predictionof stability based on accelerated studies can provide properand valuable information (Yoshioka et al., 1994; Mazzobreet al., 1997). However, real-time stability testing shouldbe conducted in parallel with accelerated studies for theselection of the optimal final formulation. In this work, theaccelerated stability study was conducted with the mainobjective of identifying the most stable antivenom formu-lation to thermal stress.

Antivenoms formulatedwith different concentrations ofsorbitol, mannitol and sucrose were subjected to an accel-erated stability test at 40 �C ± 2 �C and 75% ± 5% relativehumidity during six months. Formulations were assessedfor appearance of the cake, residual moisture content,reconstitution time, turbidity, electrophoretic pattern andneutralization of the lethal activity against the B. aspervenomat the beginning and the end of the study. Analysis ofthe antivenoms showed that, after sixmonths of incubationat 40 �C, the two formulations with sucrose and theformulation with 2% mannitol decreased their residual hu-midity (P < 0.05) as shown in Table 3. This is unexpected,because during the storage of solid proteins, there is usuallyan increase in residual moisture due to an exchange be-tween theproduct and the stopper,which is generally steamsterilized (Changet al., 2005a). Theprocesses involved in theobserved decrement of moisture should be further studied.

ulations after six months of storage at 40 �C and 75% relative humidity.

g/mL)b Residual moisture (%)a Turbidity (NTU)a,c

1.88) 5.4 ± 0.7 110.0 ± 0.01.49) NDd 50.0 ± 0.01.81) 5.6 ± 0.5 150.0 ± 1.0

2.22) 2.4 ± 0.4 78.0 ± 3.02.49) 1.8 ± 0.4 60.0 ± 0.02.11) 2.6 ± 0.2 75.0 ± 0.03.01) 2.7 ± 0.4 55.0 ± 0.03.85) 4.5 ± 0.4 120.0 ± 0.0

rvals are depicted in parenthesis.

The impact of moisture content in the storage stabilityof freeze-dried antibodies has been previously described(Breen et al., 2001; Chang et al., 2005b). It is known thatincrements in moisture decrease the Tg value, affecting thestability of the formulation (Breen et al., 2001); however,overdrying of proteins is also detrimental for stability. Inthis study the variation on moisture content during incu-bation was very small and probably does not affect thestability of formulations. Nevertheless, the appropriate re-sidual moisture and its impact in other snake antivenomscannot be extrapolated from our findings, and must beestablished for each formulation and for eachmanufacturerthrough their own stability studies.

The assessment of reconstitution time after six monthsof incubation revealed a statistically significant increase inthis parameter for all the formulations (P < 0.001). Underthe conditions of this study, we did not find any correlationbetween the excipient content and reconstitution time, andthe control antivenom had the highest reconstitution time,whereas the antivenoms freeze-dried with 5% sucrose and2%mannitol showed the smallest increase in reconstitutiontime. These results suggest that the presence of mannitol orsucrose exerts a protective effect on the mechanical struc-ture of the cake, preventing the collapse and possibly fa-voring an internal microscopic structure with many poresthat allow the hydration of the samples (Overcashier et al.,1999).

The turbidity of antivenoms at the end of stability studyshowed a significant increase for all the formulations.However, this increment was lower in case of antivenomsfreeze-dried with either 0.5 M sorbitol or 5% sucrose, andthere was a correlation between turbidity and the con-centration of the excipient. Protein turbidity has beenassociated with the formation of insoluble aggregates,

Fig. 2. SDS-PAGE of freeze-dried antivenoms at the beginning of the study and aftesamples were loaded in 7.5% polyacrylamide gels in the presence of SDS. Proteins wbeginning of the accelerated test. Lane 1: control antivenom, lane 2: antivenom fsorbitol. Panel B: Antivenoms with sorbitol after six months of accelerated test. Lanlane 3: antivenom formulated with 0.5 M sorbitol. Panel C: antivenoms with sucroselane 2: antivenom formulated with 2% sucrose; lane 3: antivenom formulated with 5respectively. Notice an intense band at the upper part of the samples that it is notmolecular mass protein aggregates. This band is present with less intensity in theantivenoms formulated with mannitol and sucrose at the beginning of the stability

which have been associated with the development ofadverse events during administration (Cromwell et al.,2006). The protective effect of polyols like sorbitol andmannitol in the development of turbidity in liquid anti-venoms has already been reported (Segura et al., 2009;Rodrigues-Silva et al., 1997, 1999). However, the use of su-crose in the stabilization of liquid or freeze-dried anti-venoms has not been previously described.

The formation of high molecular mass aggregates dur-ing the accelerated stability study was monitored bychanges in the electrophoretic profile of antivenoms(Fig. 2). Electrophoretic analysis evidenced the appearanceof an intense band at the upper part of the gel after 6months of incubation at 40 ± 2 �C, corresponding probablyto high molecular mass protein aggregates. This band wasnot present in the samples with sorbitol at the beginning ofthe study (Panel A) but it appeared in all formulations withsorbitol, mannitol and sucrose (Panels B and C), but wasless intense in the sample formulated with 5% sucrose(Panel C, lane 3). The electrophoretic profiles of antivenomsformulated with mannitol and sucrose at the beginning ofthe stability test (not shown) were similar to the onedescribed for samples with sorbitol.

These results agree with those reported by other re-searchers, in that sorbitol only slightly protected an IgG1antibody formulation against aggregation during storage,whereas sucrose improved stability significantly (Changet al., 2005a). It is postulated that an excipient systemthat remains at least partially amorphous is necessary forstabilization (Pikal et al., 1991); however, the observationthat sorbitol formulations showed poor stability regardingaggregation, demonstrates that an amorphous excipientsystem is not the only condition necessary for stability ofantivenoms.

r six months of storage at 40�± 2 �C and 75% relative humidity. Non-reducedere stained with Coomassie Brilliant Blue R-250. Panel A: antivenoms at theormulated with 0.05 M sorbitol; lane 3: antivenom formulated with 0.5 Me 1: control antivenom; lane 2: antivenom formulated with 0.05 M sorbitol;and mannitol after six months of accelerated test. Lane 1: control antivenom;% sucrose; lanes 4 and 5: antivenoms formulated with 2% and 10% mannitol,

present at the beginning of the study and that probably corresponds to highsample formulated with 5% sucrose (Panel C). The electrophoretic profile oftest was similar to the one exhibited by samples formulated with sorbitol.

The properties of the solid protein formulation play animportant role in the storage stability. It has been demon-strated that storage of samples above the glass transitiontemperature of the solid matrix (Tg), might lead to rapiddegradation (Chang et al., 2005a; Duddu and Dal Monte,1997). In this sense, instability of antivenoms formulatedwith sorbitol and mannitol during storage at 40 �C could berelated to a low glass transition temperature (Tg) for theseformulations, probably near or below 40 �C, which could bedirectly affecting the physical and chemical stability of thesamples. Nevertheless, this issue deserves furtherinvestigation.

The effect of the storage on the neutralizing efficacy ofantivenoms was assessed by determining the ED50 of theformulations. Results showed that, after six months of in-cubation, the neutralizing activity decreased significantlyfor the two formulations with mannitol.

The stabilizing effect of saccharides and polyols duringfreeze-drying and the storage of proteins has been explainedby two main mechanisms: (a) a thermodynamic stabiliza-tion through direct interactions between protein andexcipient (e.g., hydrogen bonds) that substitute surroundingwater molecules (Allison et al., 1999), and (b) a kineticmechanism that reduces chemical degradation by embed-ding the protein in a glass state of lower molecular mobility(Arakawa et al., 2001). The effectiveness of sorbitol and su-crose to protect the neutralizing activity of antivenomsduring six months of storage at 40 �C can be explained interms of these mechanisms. Moreover, the decrease in thepotency of antivenoms formulated with mannitol might berelated to the formation of a crystalline matrix of mannitol,due to the annealing phase carried out during the freeze-drying process. This relationship between the degree ofcrystallization and the decrease on the protective effect ofmannitol in protein formulations has already been reported(Izutsu et al., 1993; Pyne et al., 2003). Our results suggestthat the main mechanism involved in the stabilization offreeze-dried snake antivenoms is likely to be the formationof an amorphousmatrix instead of the substitution of water,a hypothesis that requires further studies.

The visual examination of the freeze-dried formulationsafter six months of storage at high temperature showedthat all samples retained their cake structure. The mannitolpresent in formulations probably crystallized completely inthe annealing step of the freeze-drying process and,therefore, remained stable during storage. All the anti-venoms, except the sample formulated with 5% sucrose,exhibited a color change to browning with time. Thischange in the appearance of the cake was more pro-nounced with increased concentrations of mannitol andsorbitol. Reducing sugars, such as glucose and lactose, canreact with some residues in proteins to form carbohydrateadducts via the Maillard reaction. This process is also calledbrowning reaction or glycation and has been studiedmainly in the food industry (Martins et al., 2001; Burdurluand Karadeniz, 2003). The browning of solid pharmaceu-tical proteins during storage has been observed for otherauthors (Li et al., 1996; Wu et al., 1998; Andya et al., 1999),but this phenomenon has not so far been described forfreeze-dried antivenoms. Mannitol and sorbitol are notreducing sugars; however impurities present in these

excipients can react with some residues in the proteins(Dubost et al., 1996), and this might explain the observedcolor change in antivenoms freeze-dried with these poly-ols. A direct relationship between the potency of the anti-venoms and the change in the appearance of the cake wasnot observed in our experiments. Nevertheless, the role ofexcipient impurities in promoting drug reactions in thesolid state needs to be addressed in future studies sincethese polyols are widely used in lyophilized pharmaceu-tical forms.

4. Conclusions

Successful freeze-drying of antivenoms requires theconservation of the neutralizing potency of the antibodies aswell the improvementof thephysicochemical stability of theprotein. This study investigated the use of sucrose, mannitoland sorbitol as stabilizers for snake antivenoms formulatedas freeze-driedproducts.DSCprovidesuseful informationonthe freezing behavior of different formulations and allowedthe design of an appropriate freeze drying process. After sixmonthsof storage at40 �C, all the formulationspresented thesame residual humidity content, but significant differenceswere observed in turbidity, reconstitution time and elec-trophoretic pattern. Moreover, all formulations, exceptantivenoms freeze-dried with mannitol, exhibited the samepotency for the neutralization of lethal effect of B. aspervenom.The5% sucrose formulation showed thebest stabilityamong the samples tested, while the mannitol and sorbitolformulations underwent significant destabilization andturned brown. Results indicate that sucrose couldperformasa better stabilizer than mannitol and sorbitol in the formu-lation of freeze-dried antivenoms. This formulation maybecome a good alternative for the production of more stableantivenoms in regionsof theworldwherehigh temperaturesare common and the cold chain is poor.

Ethical statement

This manuscript presents an experimental study per-formed following the standardprocedureof scientific ethics.

Acknowledgments

The authors thank our colleagues of Instituto ClodomiroPicado, as well as Sergio Ramírez for his collaboration onDSC analyses. This study was financially supported byVicerrectoría de Investigaci�on (Universidad de Costa Rica)project 741-B2-091 and by the program CYTED (projectBIOTOX 212RT0467).

Conflicts of interest

The authors declare that there are no conflicts ofinterest.

Transparency document

Transparency document related to this article can befound online at http://dx.doi.org/10.1016/j.toxicon.2014.07.015.

References

Allison, S., Chang, B., Randolph, T., Carpenter, J., 1999. Hydrogen bondingbetween sugar and protein is responsible for inhibition ofdehydration-induced protein unfolding. Arch. Biochem. Biophys. 365,289e298.

Andya, J.D., Maa, Y.F., Costantino, H.R., Nguyen, P.A., Dasovich, N.,Sweeney, T.D., Hsu, C.C., Shire, S.J., 1999. The effect of formulationexcipients on protein stability and aerosol performance of spray-driedpowders of a recombinant humanized anti-IgE monoclonal antibody.Pharm. Res. 16, 350e358.

Arakawa, T., Prestrelski, S.J., Kenney, W.C., Carpenter, J.F., 2001. Factorsaffecting short-term and long-term stabilities of proteins. Adv. DrugDeliv. Rev. 46, 307e326.

Bon, C., 1996. Serum therapy was discovered 100 years ago. In: Bon, C.,Goyffon, M. (Eds.), Envenomings and Their Treatments. EditionsFondation Marcel M�erieux, Lyon, pp. 3e9.

Breen, E.D., Curley, J.G., Overcashier, D.E., Hsu, C.C., Shire, S.J., 2001. Effectof moisture on the stability of a lyophilized humanized monoclonalantibody formulation. Pharm. Res. 18, 1345e1353.

Burdurlu, H.S., Karadeniz, F., 2003. Effect of storage on non enzymaticbrowning of apple juice concentrates. Food Chem. 80, 91e97.

Chang, L., Shepherd, D., Sun, J., Ouellette, D., Grant, K., Tang, X., Pikal, M.,2005a. Mechanism of protein stabilization by sugars during freeze-drying and storage: native structure preservation, specific interac-tion, and/or immobilization in a glassy matrix? J. Pharm. Sci. 94,1427e1444.

Chang, L., Shepherd, D., Sun, J., Tang, X., Pikal, M., 2005b. Effect of sorbitoland residual moisture on the stability of lyophilized antibodies: im-plications for the mechanism of protein stabilization in the solid state.J. Pharm. Sci. 94, 1445e1455.

Council of International Organizations of Medical Sciences (CIOMS), 1986.The International Guiding Principles for Biomedical ResearchInvolving Animals. Z. Bankowski and N. Howard-Jones, (Geneva).

Cromwell, M., Hilario, E., Jacobson, F., 2006. Protein aggregation andbioprocessing. AAPS J. 8, E572eE579.

Dubost, D., Kaufman, M., Zimmerman, J., Bogusky, M., Coddington, A.,Pitzenberger, S., 1996. Characterization of a solid state reactionproduct from a lyophilized formulation of a cyclic heptapeptide. Anovel example of an excipient-induced oxidation. Pharm. Res. 13,1811e1814.

Duddu, S., Dal Monte, P., 1997. Effect of glass transition temperature onthe stability of lyophilized formulations containing a chimeric ther-apeutic monoclonal antibody. Pharm. Res. 14, 591e595.

Gerring, D., King, T.R., Branton, R., 2013. Validating a faster method forreconstitution of crotalidae polyvalent immune fab (ovine). Toxicon69, 42e49.

Guti�errez, J.M., Le�on, G., Burnouf, T., 2011. Antivenoms for the treatmentof snakebite envenomings: the road ahead. Biologicals 39, 129e142.

Heller, M., Carpenter, J., Randolph, T., 1999. Protein formulation andlyophilization cycle design: prevention of damage due to freeze-concentration induced phase separation. Biotechnol. Bioeng. 63,166e174.

Hill, R.E., Bogdan, G.M., Dart, R.C., 2001. Time to reconstitution: purifiedFab antivenom vs unpurified IgG antivenoms. Toxicon 39, 729e731.

Imamura, K., Ogawa, T., Sakiyama, T., Nakanishi, K., 2003. Effects of typesof sugar on the stabilization of protein in the dried state. J. Pharm. Sci.92, 266e274.

Izutsu, K., Yoshioka, S., Terao, T., 1993. Decreased protein-stabilizing ef-fects of cryoprotectants due to crystallization. Pharm. Res. 10,1232e1237.

Kasper, J., Friess, W., 2011. The freezing step in lyophilization: pysico-chemical fundamentals, freezing methods and consequences onprocess performance and quality attributes of biopharmaceuticals.Eur. J. Pharm. Biopharm. 78, 248e263.

Li, S., Patapoff, T.W., Overcashier, D., Hsu, C., Nguyen, T.H., Borchardt, R.T.,1996. Effects of reducing sugars on the chemical stability of humanrelaxin in the lyophilized state. J. Pharm. Sci. 85, 873e877.

Martins, S., Jongen, W., van Boekel, M., 2001. A review of maillard reactionin food and implications to kinetic modelling. Trends Food Sci.Technol. 11, 364e373.

Maury, M., Murphy, K., Kumar, S., Mauerer, A., Lee, G., 2005. Spray-dryingof proteins: effects of sorbitol and trehalose on aggregation and FT-IRamide I spectrum of an immunoglobulin G. Eur. J. Pharm. Biopharm.59, 251e261.

Mazzobre, M.F., Buera, M.P., Chirife, J., 1997. Glass transition and termalstability of lactase in low-moisture amorphous polymeric matrices.Biotechnol. Prog. 13, 195e199.

Meister, E., Gieseler, H., 2009. Freeze-dry microscopy of protein/sugarmixtures: drying, behavior, interpretation of collapse temperaturesand a comparison to corresponding glass transition data. J. Pharm. Sci.98, 3072e3087.

Otero, R., Guti�errez, J., Mesa, M., Duque, E., Rodríguez, O., Arango, J.L.,G�omez, F., Toro, A., Cano, F., Rodríguez, L.M., Caro, E., Martínez, J.,Cornejo, W., G�omez, L.M., Uribe, F.L., C�ardenas, S., Nú~nez, V., Díaz, A.,2002. Complications of Bothrops, Porthidium and Bothriechis snake-bites in Colombia. A clinical and epidemiological study of 39 casesattended in a University Hospital. Toxicon 40, 1107e1114.

Overcashier, D., Patapoff, T., Hsu, C., 1999. Lyophilization of protein for-mulations in vials: investigation of the relationship between resis-tance to vapor flow during primary drying and small-scale productcollapse. J. Pharm. Sci. 88, 688e695.

Parvin, R., Pande, S.V., Venkitasubramanian, T.A., 1965. On the colorimetricbiuret method of protein determination. Anal. Biochem 12, 219e229.

Pikal, M.J., Shah, S., 1990. The collapse temperature in freeze-drying:dependence on measurement methodology and rate of waterremoval from the glassy phase. Int. J. Pharm. 62, 165e186.

Pikal, M.J., Dellerman, K.M., Roy, M.L., Riggin, R.M., 1991. The effects offormulation variables on the stability of freeze-dried human growthhormone. Pharm. Res. 8, 427e436.

Pyne, A., Chatterjee, K., Suryanarayanan, R., 2003. Solute crystallization inManitol-Glycine systems, implications on proteins stabilization infreeze-dried formulations. J. Pharm. Sci. 92, 2272e2283.

Quan, A., Quan, D., Curry, S., 2010. Improving crotalidae polyvalent im-mune fab reconstitution times. Am. J. Emer. Med. 28, 593e595.

Rodrigues-Silva, R., Martins, M.S., Magalhaes, A., Santoro, M.M., 1997.Purification and stability studies of immunoglobulins from Lachesismuta muta antivenom. Toxicon 35, 1229e1238.

Rodrigues-Silva, R., Antunes, G., Velarde, D.T., Santoro, M.M., 1999. Ther-mal stability of hyperimmune horse antivenoms. Toxicon 37, 33e45.

Rojas, G., Jim�enez, J.M., Guti�errez, J.M., 1994. Caprylic acid fractionation ofhyperimmune horse plasma: description of a simple procedure forantivenom production. Toxicon 32, 351e363.

Sarciaux, J.M., Mansour, S., Hageman, M.J., Nail, S.L., 1999. Effects of buffercomposition and processing conditions on aggregation of bovine IgGduring freeze-drying. J. Pharm. Sci. 88, 1354e1361.

Schersch, K., Betz, O., Garidel, P., Muehlau, S., Bassarab, S., Winter, G.,2010. Systematic investigation of the effect of lyophilizate collapse onpharmaceutically relevant proteins I: stability after freeze-drying. J.Pharm. Sci. 99, 2256e2278.

Searles, J.A., Carpenter, J.F., Randolph, T.W., 2001. Annealing to optimizethe primary drying rate, reduce freezing-induced drying rate het-erogeneity, and determine Tg' in pharmaceutical lyophilisation. J.Pharm. Sci. 90, 872e887.

Segura, A., Herrera, M., Gonz�alez, E., Vargas, M., Solano, G., Guti�errez, J.M.,Le�on, G., 2009. Stability of equine IgG antivenoms obtained y caprylicacid precipitation: towards a liquid formulation stable at tropicalroom temperature. Toxicon 53, 609e615.

Sharma, V.K., Kalonia, D.S., 2004. Effect of vacuum drying on protein-mannitol interactions: the physical state of mannitol and proteinstructure in the dried state. AAPS Pharm. Sci. Tech. 5, 1e12.

Solano, G., Segura, A., Herrera, M., G�omez, A., Villalta, M., Guti�errez, J.M.,Le�on, G., 2010. Study of the design and analytical properties of thelethality neutralization assay used to estimate antivenom potencyagainst Bothrops asper snake venom. Biologicals 38, 577e585.

Theakston, R.D.G., Warrell, D.A., Griffiths, E., 2003. Report of a WHOworkshop on the standardization and control of antivenoms. Toxicon41, 541e557.

Wang, B., Tchessalov, S., Cicerone, M., Warne, N., Pikal, M., 2009. Impact ofsucrose level on storage stability of proteins in freeze-dried solids: II.Correlation of aggregation rate with protein structure and molecularmobility. J. Pharm. Sci. 98, 3145e3166.

Wang, W., 2000. Lyophilization and development of solid protein phar-maceuticals. Int. J. Pharm. 203, 1e60.

Wu, S., Leung, D., Vemuri, S., Shah, D., Yang, B., Wang, J., 1998. Degradationmechanism of lyophilized basic fibroblast growth factor (BFGF) pro-tein in sugar formulation. Pharm. Sci. (Suppl. l) S-539.

Yoshioka, S., Aso, Y., Izutsu, K., Kojima, S., 1994. Is stability predictionpossible for protein drugs? Denaturation kinetics of b- Galactosidasein solution. Pharm. Res. 11, 1721e1725.

Freeze-dried snake antivenoms formulated with sorbitol, sucrose or mannitol: Comparison of their...

Documents