Ecv

Ma

b

a

ARRAA

KSCPIPV

1

uptsn21ocw

flaee

h0

Carbohydrate Polymers 152 (2016) 12–18

Contents lists available at ScienceDirect

Carbohydrate Polymers

j ourna l ho me pa g e: www.elsev ier .com/ locate /carbpol

ffects of solution conditions on characteristics and size exclusionhromatography of pneumococcal polysaccharides and conjugateaccines

ahsa Hadidia, John J. Buckleyb, Andrew L. Zydneya,∗

Department of Chemical Engineering, The Pennsylvania State University, University Park, PA 16802, United StatesPfizer Inc., 700 Chesterfield Village Parkway, Chesterfield, MO 63017, United States

r t i c l e i n f o

rticle history:eceived 13 May 2016eceived in revised form 23 June 2016ccepted 24 June 2016vailable online 25 June 2016

eywords:ize exclusion chromatography

a b s t r a c t

Molecular properties of bacterial polysaccharides and protein-polysaccharide conjugates play animportant role in the efficiency and immunogenicity of the final vaccine product. Size exclusion chro-matography (SEC) is commonly used to analyze and characterize biopolymers, including capsularpolysaccharides. The objective of this work was to determine the effects of solution ionic strength andpH on the SEC retention of several capsular polysaccharides from S. pneumoniae bacteria in their nativeand conjugated forms. The retention time of the charged polysaccharides increased with increasing ionicstrength and decreasing pH due to compaction of the polysaccharides associated with a reduction in the

apsular polysacchariderotein-polysaccharide conjugatesonic strengthneumococcusaccines

intramolecular electrostatic interactions. The calculated radius of gyration was in good agreement withmodel calculations based on the worm-like chain model accounting for the increase in chain stiffness andexcluded volume of the charged polysaccharide at low ionic strength. These results provide importantinsights into the effects of solution ionic strength on physical properties and SEC behavior of capsularpolysaccharides and their corresponding conjugates.

© 2016 Elsevier Ltd. All rights reserved.

. Introduction

Capsular polysaccharides from pathogenic bacteria have beensed to produce vaccines against important diseases such asneumonia and meningitis. The efficiency and immunogenicity ofhe polysaccharide and protein-polysaccharide conjugate vaccinesignificantly depends on the molecular properties of the compo-ents such as molecular weight and/or size (Astronomo & Burton,010; Jennings, 1990; Jennings, Rosell, Katzenellenbogen, & Kasper,983). For example, Lee et al. (2009) evaluated the effect of sizen the immunogenicity of meningococcal polysaccharide-proteinonjugates, and clearly demonstrated that the larger conjugatesere more immunogenic.

Size exclusion chromatography (SEC) has been used extensivelyor characterization of carbohydrate polymers, including capsu-ar bacterial polysaccharides (and their conjugates) that are used

s important vaccine products to ensure both the stability andffective immunogenicity (Gaborieau & Castignolles, 2011; Leet al., 2009; Mori & Barth, 2013; Pato, Antonio de Pádua, & da

∗ Corresponding author.E-mail address: zydney@engr.psu.edu (A.L. Zydney).

ttp://dx.doi.org/10.1016/j.carbpol.2016.06.095144-8617/© 2016 Elsevier Ltd. All rights reserved.

Silva Junior, 2006; Von Hunolstein, Parisi & Bottaro, 2003; Yu &Rollings, 1987). For example, Lee et al. (2009) used size exclusionchromatography for the purification and analysis of a meningococ-cal conjugated vaccine. Hunolstein and colleagues (Parisi & VonHunolstein, 1999; Von Hunolstein et al., 2003; Von Hunolstein,Parisi, & Recchia, 1999) used the PL Aquagel-OH 60 column to deter-mine the molecular size distribution of high-molecular-weightHaemophilus influenzae type b-tetanus toxoid conjugate vaccines.Jumel, Ho, and Bolgiano (2002) evaluated the mean molecular massof three different protein-meningococcal C conjugates and theirconstituent proteins using a Superose 6 size exclusion chromatog-raphy column. Harding et al. (2012) comprehensively characterizeda wide range of native capsular polysaccharides from Streptococ-cus pneumoniae using size exclusion chromatography coupled withmulti-angle laser light scattering (MALLS). Thiébaud et al. (2014)recently discussed the development and validation of a high per-formance SEC method to characterize Haemophilus influenzae typeb polysaccharides and their corresponding conjugates, while Hoet al. (2002) used SEC to track the thermal stability of tetanus tox-

oid conjugate vaccines, showing an increase in the relative amountof a low molecular weight component after storage at elevatedtemperatures.

M. Hadidi et al. / Carbohydrate Polymers 152 (2016) 12–18 13

Table 1Composition of buffers used in SEC studies at different pH.

pH Composition

7 KCl (or CaCl2/MgSO4) buffered with Bis-Tris

bhacoanc(siim

tstdt

2

lsOomdfbbi(wwsa(sd

aaIC(swTm2t

SMt

Table 2Properties of dextran standards (data from manufacturer).

Average Molecular weight (Da) Mw/Mn

Mw Mp Mn

DXT165 165,500 150,000 110,800 1.49DXT325 326,600 245,000 205,000 1.59DXT550 548,300 350,000 346,500 1.58DXT750 749,500 560,000 500,500 1.50DXT1300 1,360,000 1,199,000 877,800 1.55DXT2800 2,800,000 2,655,000 1,970,000 1.42

Mw = weight average molecular weight.Mp = peak molecular weight.Mn = number average molecular weight.

6 KCl buffered with Bis-Tris5 KCl buffered with sodium acetate

However, essentially all of these SEC studies examined theehavior of the polysaccharides and their conjugates at relativelyigh ionic strength (greater than 0.1 M) and near neutral pH. Bednarnd Hennessey (1993) noted that the elution volume for pneumo-occal capsular polysaccharides at pH 7.2 was largely independentf the ionic strength of the mobile phase from 0.05 to 0.5 M,lthough no details were provided on the actual profiles. In contrast,umerous studies have demonstrated the importance of solutiononditions on SEC retention for charged macromolecules like DNALatulippe & Zydney, 2009), proteins (Pujar & Zydney, 1998), andynthetic polymers (Herman, Field, & Abbott, 1981). These changesn retention are usually attributed to a combination of intra- andnter-molecular electrostatic interactions affecting both macro-

olecular size and partitioning.The objective of this study was to obtain quantitative data for

he SEC behavior of the native and conjugated forms of severalerotypes of S. pneumoniae capsular polysaccharides. SEC reten-ion data were compared with that of narrow molecular weightextran standards, with the results analyzed in terms of availableheoretical models for the radius of gyration of charged polymers.

. Materials and methods

Size exclusion chromatography was performed using an Agi-ent 1200 series high performance liquid chromatography (HPLC)ystem (Agilent Technologies, CA) equipped with a PL Aquagel-H 60 size exclusion column (7.5 × 300 mm) having a particle sizef 8 �m (Agilent Technologies, CA). The PL Aquagel-OH is a rigidacroporous copolymer resin with a very hydrophilic polyhy-

roxyl surface capable of resolving species with molecular weightsrom 200 to 10,000 kDa. The neutral hydrophilic surface minimizesoth electrostatic and hydrophobic interactions; this column haseen widely used for the analysis of water soluble polymers includ-

ng high molecular weight polysaccharides and vaccine conjugatesParisi & Von Hunolstein, 1999). The flow rate of the running bufferas maintained at 0.8 mL/min and the sample injection volumeas 80 �L. Sample detection was performed using an Agilent 1200

eries refractive index detector (RID) at 35 ◦C (maintained using column oven). Data were analyzed using ChemStation softwareAgilent). Limited data were obtained with a Malvern OMNISECystem (Westborough, MA) with a multi-angle light scatteringetector with 20 detectors arranged from 12 to 164◦.

Buffer solutions were prepared by dissolving pre-weighedmounts of appropriate salts in deionized water obtained from

NANOpure® Diamond water purification system (Barnsteadnternational) with conductivity less than 56 nS/cm. KCl (BDHhemicals, BDH0258), CaCl2 (EMD Chemicals, CX0156-1), MgSO4BDH Chemicals, BDH9246), Bis-Tris (MP Biomedical, 101038),odium acetate (Sigma, S7670), and HCl (EM Science, HX0603-3)ere used to adjust the ionic strength and pH as listed in Table 1.

he solution pH was measured using a Thermo Orion pH meter. Theobile phase was prefiltered through a 0.2 �m pore size Supor®

00 membrane (Pall Corporation) prior to use to remove any par-icles or undissolved salts.

Solutions of the native and conjugated polysaccharides from. pneumoniae bacteria were provided by Pfizer, Inc. (Chesterfield,O). The conjugates were formed by coupling an activated form of

he capsular polysaccharide, prepared by partial hydrolysis using

Fig. 1. Elution volume in size exclusion chromatography as a function of solutionionic strength for 0.1 g/L solutions of Serotype A, B, and C polysaccharides at pH 7.

periodate, to CRM197, a non-toxic cross-reacting mutant of thediphtheria toxin protein (Biemans, Duvivier, Gavard, & Poolman,2010). SEC samples were prepared by diluting with the desiredmobile phase to obtain a concentration of approximately 0.1 g/L.The resulting solutions were filtered through 0.22 �m Acrodisc®

syringe filters (Pall Corporation) immediately prior to use. Thepolysaccharide solutions were stored at 4◦ C (to minimize anydegradation of the samples) and gradually brought to room temper-ature (22 ± 3◦ C) prior to each experiment. Data were also obtainedwith dextran standards obtained from American Polymer Stan-dards Corp. (Mentor, OH) as listed in Table 2. Total void volume (Vt)of the column was determined using d-glucose (Sigma, G5767) withMw = 180 Da as a completely accessible small molecule while theinterparticle void volume (Vo) was evaluated using lambda DNA(New England Biolabs, N3011S) with Mw = 3.2 × 107 Da, whichis 3.2 times larger than the reported exclusion limit for the PLAquagel-OH 60 column.

3. Results and discussion

3.1. Effect of ionic strength on polysaccharides

Fig. 1 shows data for the elution (or retention) volume of

three polysaccharide serotypes (properties in Table 3) using the PLAquagel-OH 60 size exclusion column with 10 mM Bis-Tris at pH 7plus added KCl (with molar concentrations between 5 and 250 mM)as the running buffer. Data obtained with different polysaccharide

14 M. Hadidi et al. / Carbohydrate Polymers 152 (2016) 12–18

Table 3Properties of polysaccharide serotypes.

Serotype Molecular weight (kDa) Charge Density Charge Parameter, � Repeating Unit Length (nm) Contour Length, L (nm)

Serotype A 1940 1/2 0.894 0.8 430258 2 1520

1.6 ND

N

ctwcD

I

wrmiabsnsci

ncsivcrc

R

wR

�

wmnNti

wtllrwAtiorTSpt

Serotype B 710 1/5 0.3Serotype C ND 0 0

D = not determined.

oncentrations (from 0.1 to 1 g/L) gave identical results for bothhe retention volumes/times and peak shape. The total void volumeas estimated using glucose as 9.91 ± 0.08 mL while the interparti-

le volume was evaluated as 6.40 ± 0.12 mL using results for lambdaNA. Data were plotted as a function of the buffer ionic strength:

= 12

∑i

z2i C2

i (1)

here zi and Ci are the net charge and concentration of each ion,espectively, with the concentration of the ionized Bis-Tris deter-ined based on its pKa. The elution volume for Serotype C is

ndependent of the ionic strength, consistent with the absence ofny charged groups in its structure. In contrast, Serotypes A and Both show a sharp increase in elution volume with increasing ionictrength, corresponding to a reduction in their effective hydrody-amic volume. The net result is that Serotypes B and C have fairlyimilar retention volumes at high ionic strength but have dramati-ally different elution behavior at low ionic strength. This behaviors discussed in more detail subsequently.

The data in Fig. 1 were used to evaluate the effective hydrody-amic radius of the polysaccharides based on a calibration curveonstructed with the series of relatively monodisperse dextrantandards (Table 2). The dextrans were run at each ionic strength,mmediately before/after the polysaccharides; the dextran elutionolume was nearly independent of ionic strength (and pH, dis-ussed in next section), with variations less than ±1%. The dextranadius (at each ionic strength) was calculated from the empiricalorrelation given by Granath (1958):

SEC = 3.1 × 10−11(Mw)0.47752 (2)

here the radius is in meters and the molecular weight is in Da.esults are plotted in Fig. 2 as a function of the Debye length:

D =√

εkBT

2NAe2I(3)

here ε is the electrical permittivity of the solution, kB is the Boltz-ann constant, T is the absolute temperature, NA is Avogadro’s

umber, e is the electron charge, and I is the solution ionic strength.ote that Fig. 2 is equivalent to plotting the data versus I-1/2 due to

he dependence of the Debye length on the ionic strength as givenn Eq. (3).

The effective radius of Serotype A in the 250 mM KCl solutionas 52 nm, with RSEC ≈ 43 nm by extrapolation of the data in Fig. 2

o zero Debye length (infinite ionic strength). This value is in excel-ent agreement with the 44 nm radius determined by multi-angleight scattering (MALS) using the Malvern OMNISEC. The MALSadius of Serotype B was 29 nm, which is also in good agreementith the SEC data in Fig. 2. The hydrodynamic radii of the Serotype

and B polysaccharides increase by 80 nm and 42 nm, respec-ively, as the Debye length increases from 0.61 to 4.3 nm. This largencrease in polysaccharide radius is due primarily to the expansionf the polysaccharide associated with intramolecular electrostaticepulsion between the negatively charged carboxylic acid groups.

he greater effect of ionic strength on the hydrodynamic radius oferotype A is consistent with its higher charge density. The SECartitioning behavior can also be affected by intermolecular elec-rostatic interactions associated with the distortion of the external

Fig. 2. Effective hydrodynamic radius determined by size exclusion chromatogra-phy using dextran standards for 0.1 g/L solutions of Serotype A and B polysaccharidesat pH 7 as a function of Debye length.

electrical double layer surrounding the polysaccharide. However,previous studies have shown that this latter effect is relatively small(Pujar & Zydney, 1998); this behavior was confirmed using SECdata for the globular protein thyroglobulin (Mw = 660 kDa), whichshowed less than a 12 nm increase in size over the same range ofionic strength (data not shown).

Additional insights into the effects of intramolecular inter-actions on the effective hydrodynamic radius were obtainedby performing experiments in which the ionic strength wasadjusted using CaCl2 instead of KCl. The hydrodynamic radius againincreased with increasing Debye length, but the dependence wasmuch weaker than that seen in the KCl solutions. For example, thehydrodynamic radius of Serotype A at a Debye length of 2.2 nm,corresponding to an ionic strength of 20 mM, is only 67 nm inCaCl2 compared to 93 nm in KCl (both with 10 mM Bis-Tris asthe buffer). This is likely due to the ability of the divalent Ca2+

to form a salt bridge between the negatively charged carboxylicacid groups, further reducing the expansion of the polysaccha-ride (Latulippe & Zydney, 2008; Medina-Torres, Brito-De La Fuente,Torrestiana-Sanchez, & Katthain, 2000). Data obtained using MgSO4(divalent–divalent salt) were nearly indistinguishable from thedata with CaCl2 at the same ionic strength (i.e., at the same Debyelength).

3.2. Effect of pH on polysaccharides

The effect of solution pH on the effective hydrodynamic radius ofthe polysaccharides is shown in Fig. 3. The effective radius of neutral

Serotype C was independent of pH (RSEC = 24 ± 1.2 nm) as expected(data not shown). In contrast, the hydrodynamic radius of SerotypesA and B decreased with decreasing solution pH, particularly inthe low ionic strength solution, consistent with a reduction in the

M. Hadidi et al. / Carbohydrate Polymers 152 (2016) 12–18 15

Fpo

deArrsia

paulwpnetdt

3

etV

R

tac

R

tei

ig. 3. Effective hydrodynamic radius determined by size exclusion chromatogra-hy as a function of pH at solution ionic strength of 5 and 100 mM for 0.1 g/L solutionsf Serotype A and B polysaccharides.

egree of ionization of the carboxylic acid groups (pKa ≈ 4.7). Thisffect was most pronounced for the more highly charged Serotype, with the radius decreasing from 132 to 53 nm as the pH waseduced from pH 7 to 5 in the 5 mM ionic strength solution. The netesult is that Serotype A (effective radius of 53 nm) has a smallerize than Serotype B (effective radius of 69 nm) at pH 5 and lowonic strength, with the reverse behavior seen at both higher pHnd higher ionic strength.

The large change in effective size of Serotype A with solutionH is likely due to the effects of charge regulation. The carboxyliccid groups in both Serotypes A and B should have similar pKa val-es. However, the high linear charge density of Serotype A will

ead to a strong partitioning of positive H+ ions into the void spaceithin the polysaccharide chain. This will lead to a decrease in localH = −log[H+] and in turn an increase in the protonation of theegatively-charged carboxylic acids, reducing the intramolecularlectrostatic repulsion and effective size of the polysaccharide. Thisype of charge regulation will be less pronounced with Serotype Bue to the lower linear charge density and thus the weaker parti-ioning of H+.

.3. Theoretical analysis

The radius of gyration of an uncharged polysaccharide can bevaluated using the worm-like chain model in terms of the persis-ence length (Lp) and the contour length (L) as (De Nooy, Besemer,an Bekkum, Van Dijk, & Smit, 1996):

2g = LLp

3− L2

p + 2L3p

L−

2L4p

[1 − exp

(− L

Lp

)]L2

(4)

The persistence length provides a measure of the conforma-ional flexibility of the polymer and has a limiting value of 0 for

truly random coil and ∞ for a perfectly stiff rod. For long polymerhains, i.e., for L � Lp, Eq. (4) reduces to:

2g = LLp

3(5)

Electrostatic interactions have two primary effects on the struc-ure/size of the polysaccharides: (1) there is an increase in theffective stiffness of the polymer chain, reflected in an increasen the persistence length at low ionic strength, and (2) there is an

Fig. 4. Total persistence length vs the product of the linear charge density and theDebye length for Serotypes A, B, and C evaluated from SEC data obtained at pH 7with 0.1 g/L solutions using Eqs. (5) to (8).

increase in the excluded volume of the polymer chain with decreas-ing ionic strength. A large number of theoretical models have beendeveloped to describe both of these phenomena (Ha & Thirumalai,1995; Odijk, 1977; Skolnick & Fixman, 1977), and there is still con-siderable debate over the most appropriate theoretical framework.The simplest approach is to define an effective persistence length(Lp) as (Reed, Ghosh, Medjahdi, & Francois, 1991):

Lp = Lo + LE (6)

where LE accounts for the increase in chain stiffness due tointramolecular electrostatic interactions and Lo is the intrinsic per-sistence length for the uncharged polymer, i.e., the persistencelength in the limit of very high solution ionic strength where elec-trostatic interactions are negligible.

Several authors have shown that the electrostatic contributionto the apparent persistence length scales linearly with the Debyelength and the linear charge density (De Nooy et al., 1996; Fisher,Sochor, & Tan, 1977; Reed & Schmitz, 1994):

LE = ˇ��D (7)

where � is a proportionality constant and � is the number of chargegroups per Bjerrum length (the distance at which the electrostaticinteraction energy between two elementary charges is approxi-mately equal to the thermal energy, kT):

�B = e2

4�εoεrkBT(8)

where �B = 0.71 nm in aqueous solution.The calculated values of the apparent (total) persistence length

for Serotypes A, B, and C, calculated directly from Eq. (5) using thecontour lengths given in Table 3, are plotted in Fig. 4 as a functionof the charge parameter, ��D. In each case, the contour length wasevaluated assuming that the length of each saccharide monomerwas 0.4 nm (Hannoun & Stephanopoulos, 1986; Sabek et al., 2013).The Lp data for Serotype A show a highly linear dependence onDebye length (R2 > 0.99); the results for Serotype B show some-

what greater scatter, although the R2 value is still greater than 0.97.The different slopes for the two serotypes could be due to errorsin the calculated values of the contour length arising from uncer-tainties in the molecular weight of the polysaccharides (including

1 rate Polymers 152 (2016) 12–18

toitemlivfos

erftiDi

R

wb

a

i

Z

wi

ˇ

wS

d

E(

L

piacdSg

3c

ctfbpt

Fig. 5. Experimental radius of gyration versus the theoretical radius of gyration forpolysaccharide serotypes A, B, and C. Experimental values were determined using0.1 g/L solutions of the polysaccharides at pH 7. Theoretical values were determinedusing Eqs. (5), (6), and (9)–(14).

6 M. Hadidi et al. / Carbohyd

he presence of a molecular weight distribution). The best fit valuef the slope for Serotype A gives � = 3.2 while that for Serotype Bs � = 7.2; these values are in fairly good agreement with correla-ions presented previously for charged pullulans ( ̌ = 3.7) (De Nooyt al., 1996) and for charged polyacrylamide-polyacrylate poly-ers ( ̌ = 3.6) (Reed et al., 1991). Note that the intrinsic persistence

ength can be evaluated from extrapolation of Lp to ��D = 0, yield-ng a value between 0.25 and 1 nm. This is much smaller than thealue of Lp = 6 nm reported by Harding et al. (2012) based on dataor the intrinsic viscosity and sedimentation coefficient of a rangef polysaccharides from S. pneumoniae at pH 6.8 and 0.1 M ionictrength.

Although Eqs. (6) and (7) provide a convenient approach forvaluating the apparent (total) persistence length, and in turn theadius of gyration using Eq. (5), there is currently no direct approachor evaluating the parameter � based solely on the properties ofhe polymer/polysaccharide. An alternative approach for evaluat-ng the effective radius of a charged polysaccharide is presented bye Nooy et al. (1996). In this case, the excluded volume contribution

s evaluated explicitly as:

g2 = as

2Rgo2 (9)

here Rgo is given by Eq. (5) and as is an expansion parameter giveny the Yamakawa-Tanaka equation (Yamakawa, 1971):

s2 = 0.541 + 0.459(1 + 6.04Z)0.46 (10)

At moderate ionic strength, the excluded volume contributions dominated by electrostatic interactions, with

= 33/2

32�3/2ˇelL

1/2Lp−7/2 (11)

here ˇel is the effective volume excluded from one Kuhn segmentn the wormlike chain model:

el = 2�L2pdeff (12)

ith deff the effective diameter of the polymer chain (Fixman &kolnick, 1978):

eff = �D

[−Ln

(�B

�D

)+ 2 ln

(�)

+ 2.61]

(13)

The total persistence length Lp in Eqs. (11) and (12) is given byq. (6) with the electrostatic contribution evaluated explicitly asOdijk, 1977; Skolnick & Fixman, 1977):

E = �2�2D

4�B(14)

Note that LE given by Eq. (14) scales with the square of the chargearameter. Eq. (13) gives small (or negative) values of deff at high

onic strength; thus, the minimum value of deff was taken as 0.6 nms discussed by De Nooy et al. (1996). The model calculations areompared with the experimental data in Fig. 5. The model pre-ictions are in good agreement with the data for polysaccharideerotypes B and C, but the model clearly over-predicts the radius ofyration for Serotype A. The reason for this discrepancy is unclear.

.4. Effect of ionic strength and pH on protein-polysaccharideonjugates

Corresponding data for the effective hydrodynamic radius of theonjugates for Serotypes A, B, and C, again evaluated directly fromhe measured SEC retention volume using the dextran standards

or calibration, are shown in Fig. 6. The conjugates were producedy reacting an activated (partially hydrolyzed) version of the nativeolysaccharide with the small protein CRM197. The effective radii ofhe conjugates increase with increasing Debye length (decreasing

Fig. 6. Effective hydrodynamic radius determined by size exclusion chromatogra-phy of 0.1 g/L solutions of Conjugates A, B, and C at pH 7 as a function of Debyelength.

ionic strength), similar to the behavior seen with the free polysac-charides. For example, the hydrodynamic radius of Conjugate Aincreases from 12 to 38 nm as the ionic strength is reduced from250 to 5 mM; this compares to an increase from 52 to 132 nm for thenative serotype. The effective size of Conjugate C increases slightlywith decreasing ionic strength (from 14 to 17 nm), even thoughSerotype C is electrically neutral; this is likely due to the small netcharge on the CRM197 (isoelectric point of around pH 5.5). The netresult is that Conjugate C has the largest size at high ionic strengthbut is much smaller than Serotypes A and B at low ionic strength

(large Debye length).

The effect of solution pH on the effective hydrodynamic radiusof the conjugates is shown in Fig. 7. Experiments were only per-formed at pH 6 and 7 due to the instability of the conjugates at

M. Hadidi et al. / Carbohydrate Po

Fps

lwiaaCeF3s

4

oapibiiwtp

lptsafeeeplralf

ig. 7. Effective hydrodynamic radius determined by size exclusion chromatogra-hy as a function of pH at solution ionic strength of 5 and 100 mM using 0.1 g/Lolutions of Conjugates A, B, and C.

ower pH. In all cases, the hydrodynamic radius of the conjugateas smaller at pH 6 than at pH 7, with this difference being greater

n the low ionic strength solution. This behavior is consistent with reduction in the intramolecular electrostatic interactions associ-ted with a decrease in the net charge of the polysaccharide and theRM197 as one moves towards the isoelectric point. However, theffect of solution pH on the effective radius was relatively small.or example, the radius of Conjugate A only decreased from 38 to5 nm as the pH was reduced from pH 7 to 6 in the 5 mM ionictrength solution.

. Conclusions

The experimental results presented in this study provide somef the first quantitative data for the effects of solution ionic strengthnd pH on the effective radius of capsular polysaccharides andolysaccharide-protein conjugates, both of which are of interest

n the production of monovalent and multivalent vaccines againstacterial infections. The effective size increased significantly at low

onic strength due to the expansion of the polysaccharide due tontramolecular electrostatic interactions. Similar effects were seen

ith the conjugates, including the conjugate formed from the neu-ral Serotype C, with the electrostatic interactions now due to theresence of the charge groups on the protein.

The experimental data for the native polysaccharides were ana-yzed using available models for the radius of gyration of chargedolymers accounting for both the change in persistence length andhe excluded volume associated with the intramolecular electro-tatic interactions. The model calculations were in good qualitativegreement with the experimental results, providing a frameworkor analysis of the SEC behavior of charged polysaccharides in differ-nt solution conditions. Interestingly, the coefficient describing thelectrostatic persistence length in Eq. (7) was very similar to thatvaluated previously for charged pullulans and polyacrylamide-olyacrylate polymers, suggesting that this parameter may be

argely independent of the detailed structure of the polysaccha-

ide. It was not possible to apply this theoretical framework to thenalysis of the polysaccharide-protein conjugates since the worm-ike chain model does not provide an appropriate physical modelor the structure of the conjugates. Additional work will be required

lymers 152 (2016) 12–18 17

to fully elucidate the structure and charge characteristics of thesecomplex polysaccharide-protein conjugates.

Acknowledgement

The authors would like to acknowledge Pfizer Inc. for theirfinancial support and for providing the purified polysaccha-rides/conjugates.

References

Astronomo, R. D., & Burton, D. R. (2010). Carbohydrate vaccines: developing sweetsolutions to sticky situations? Nature Reviews Drug Discovery, 9(4), 308–324.

Bednar, B., & Hennessey, J. P. (1993). Molecular size analysis of capsularpolysaccharide preparations from Streptococcus pneumoniae. CarbohydrateResearch, 243(1), 115–130.

Biemans, R. L., Duvivier, P., Gavard, O. F. N., & Poolman, J. (2010). Immunogeniccomposition comprising S. pneumoniae polysaccharides conjugated to carrierproteins. US Patent 20120321658 A1.

De Nooy, A., Besemer, A., Van Bekkum, H., Van Dijk, J., & Smit, J. (1996).TEMPO-mediated oxidation of pullulan and influence of ionic strength andlinear charge density on the dimensions of the obtained polyelectrolyte chains.Macromolecules, 29(20), 6541–6547.

Fisher, L., Sochor, A., & Tan, J. (1977). Chain characteristics of poly(2-acrylamido-2-methylpropanesulfonate) polymers. 2. Comparison ofunperturbed dimensions and persistence lengths. Macromolecules, 10(5),955–959.

Fixman, M., & Skolnick, J. (1978). Polyelectrolyte excluded volume paradox.Macromolecules, 11(5), 863–867.

Gaborieau, M., & Castignolles, P. (2011). Size-exclusion chromatography (SEC) ofbranched polymers and polysaccharides. Analytical and Bioanalytical Chemistry,399(4), 1413–1423.

Granath, K. A. (1958). Solution properties of branched dextrans. Journal of ColloidScience, 13(4), 308–328.

Ha, B.-Y., & Thirumalai, D. (1995). Electrostatic persistence length of apolyelectrolyte chain. Macromolecules, 28(2), 577–581.

Hannoun, B. J., & Stephanopoulos, G. (1986). Diffusion coefficients of glucose andethanol in cell-free and cell-occupied calcium alginate membranes.Biotechnology and Bioengineering, 28(6), 829–835.

Harding, S. E., Abdelhameed, A. S., Morris, G. A., Adams, G., Laloux, O., Cerny, L.,et al. (2012). Solution properties of capsular polysaccharides fromStreptococcus pneumoniae. Carbohydrate Polymers, 90(1), 237–242.

Herman, D., Field, L., & Abbott, S. (1981). The size-exclusion chromatographicbehavior of synthetic water-soluble polymers on diol bonded phase supports.Journal of Chromatographic Science, 19(9), 470–476.

Ho, M. M., Mawas, F., Bolgiano, B., Lemercinier, X., Crane, D. T., Huskisson, R., et al.(2002). Physico-chemical and immunological examination of the thermalstability of tetanus toxoid conjugate vaccines. Vaccine, 20(29), 3509–3522.

Jennings, H. J., Rosell, K., Katzenellenbogen, E., & Kasper, D. (1983). Structuraldetermination of the capsular polysaccharide antigen of type II group BStreptococcus. Journal of Biological Chemistry, 258(3), 1793–1798.

Jennings, H. (1990). Capsular polysaccharides as vaccine candidates. In Bacterialcapsules. pp. 97–127. Springer.

Jumel, K., Ho, M. M., & Bolgiano, B. (2002). Evaluation of meningococcal Coligosaccharide conjugate vaccines by size-exclusionchromatography/multi-angle laser light scattering. Biotechnology and AppliedBiochemistry, 36(3), 219–226.

Latulippe, D. R., & Zydney, A. L. (2008). Salt-induced changes in plasmid DNAtransmission through ultrafiltration membranes. Biotechnology andBioengineering, 99(2), 390–398.

Latulippe, D. R., & Zydney, A. L. (2009). Size exclusion chromatography of plasmidDNA isoforms. Journal of Chromatography A, 1216(35), 6295–6302.

Lee, C.-H., Kuo, W.-C., Beri, S., Kapre, S., Joshi, J. S., Bouveret, N., et al. (2009).Preparation and characterization of an immunogenic meningococcal group Aconjugate vaccine for use in Africa. Vaccine, 27(5), 726–732.

Medina-Torres, L., Brito-De La Fuente, E., Torrestiana-Sanchez, B., & Katthain, R.(2000). Rheological properties of the mucilage gum (Opuntia ficus indica). FoodHydrocolloids, 14(5), 417–424.

Mori, S., & Barth, H. G. (2013). Size exclusion chromatography. Springer Science &Business Media.

Odijk, T. (1977). Polyelectrolytes near the rod limit. Journal of Polymer Science:Polymer Physics Edition, 15(3), 477–483.

Parisi, L., & Von Hunolstein, C. (1999). Determination of the molecular sizedistribution of Haemophilus influenzae type b–tetanus toxoid conjugatevaccines by size-exclusion chromatography. Journal of Chromatography A,847(1), 209–211.

Pato, T. P., Antonio de Pádua, R. B., & da Silva Junior, J. G. (2006). Purification of

capsular polysaccharide from Neisseria meningitidis serogroup C by liquidchromatography. Journal of Chromatography B, 832(2), 262–267.

Pujar, N. S., & Zydney, A. L. (1998). Electrostatic effects on protein partitioning insize-exclusion chromatography and membrane ultrafiltration. Journal ofChromatography A, 796(2), 229–238.

1 rate P

R

R

S

S

T

Yamakawa, H. (1971). Modern theory of polymer solutions. Harper & Row.Yu, L. P., & Rollings, J. (1987). Low-angle laser light scattering–aqueous size

exclusion chromatography of polysaccharides: molecular weight distribution

8 M. Hadidi et al. / Carbohyd

eed, W., & Schmitz, K. (1994). Macroion characterization. ACS Symposium Series,548.

eed, W. F., Ghosh, S., Medjahdi, G., & Francois, J. (1991). Dependence ofpolyelectrolyte apparent persistence lengths, viscosity, and diffusion on ionicstrength and linear charge density. Macromolecules, 24(23), 6189–6198.

abek, O. M., Ferrati, S., Fraga, D. W., Sih, J., Zabre, E. V., Fine, D. H., et al. (2013).Characterization of a nanogland for the autotransplantation of humanpancreatic islets. Lab on a Chip, 13(18), 3675–3688.

kolnick, J., & Fixman, M. (1977). Electrostatic persistence length of a wormlikepolyelectrolyte. Macromolecules, 10(5), 944–948.

hiébaud, J., Fanget, I., Jaudinaud, I., Fourrichon, L., Sabouraud, A., Talaga, P., et al.(2014). Development and validation of high-performance size exclusionchromatography methods to determine molecular size parameters ofHaemophilus influenzae type b polysaccharides and conjugates. AnalyticalBiochemistry, 453, 22–28.

olymers 152 (2016) 12–18

Von Hunolstein, C., Parisi, L., & Bottaro, D. (2003). Simple and rapid technique formonitoring the quality of meningococcal polysaccharides by high performancesize-exclusion chromatography. Journal of Biochemical and Biophysical Methods,56(1), 291–296.

Von Hunolstein, C., Parisi, L., & Recchia, S. (1999). A routine high-performancesize-exclusion chromatography to determine molecular size distribution ofHaemophilus influenzae type b conjugate vaccines. Vaccine, 17(2), 118–125.

and polymer branching determination. Journal of Applied Polymer Science,33(6), 1909–1921.

本文献由“学霸图书馆-文献云下载”收集自网络，仅供学习交流使用。

学霸图书馆（www.xuebalib.com）是一个“整合众多图书馆数据库资源，

提供一站式文献检索和下载服务”的24 小时在线不限IP

图书馆。

图书馆致力于便利、促进学习与科研，提供最强文献下载服务。

图书馆导航：

图书馆首页 文献云下载 图书馆入口 外文数据库大全 疑难文献辅助工具