TO DOWNLOAD A COPY OF THIS POSTER, VISIT WWW.WATERS.COM/POSTERS

Edouard S. P. Bouvier, Susan Serpa, Nadya Brady, Kevin D. Wyndham, Nicole L. Lawrence, Paula Hong, Uwe D. Neue, Thomas H. WalterWaters Corporation, 5 Technology Drive, Milford, MA 01757

BIOMOLECULE SEPARATIONS BY UPLC-SEC

Size exclusion chromatography is one of the primary analytical techniques used today to provide a rapid and simple means to assess protein aggregation. Aggregate impurities are carefully monitored throughout development of a protein therapeutic be-cause of their correlation to inducing immunogenicity.

SEC separations typically require long analysis times. This is primarily because the slow diffusion of large molecular weight ana-lytes within the porous media requires operation at low linear velocities to achieve optimum resolution. Mass transfer can be enhanced greatly by using smaller chromatographic particles, resulting in benefits in speed and resolution. We have recently expanded the scope of the ethylene-bridged hybrid (BEH) technology by re-engineering our particle-making process to create porous particles with almost a two-fold increase in pore volume compared with the original BEH particle. This newly designed chromatographic media (patent pending) has been shown to maintain the mechanical strength requirements necessary of a 1.7 micron particle for UPLC applications.

Another important attribute of SEC packing materials is that they interact minimally with the analytes. For traditional silica-based SEC materials, a high ionic strength mobile phase is typically employed to mitigate silanol interactions that may result in adsorption. The reduced silanol acidity on BEH particles compared to traditional silica results in decreased secondary interac-tions for charged analytes.

This presentation will discuss the benefits in speed and resolution that can be obtained with UPLC-SEC for biomolecule separa-tions. The effect of particle size, pore volume and pore size distribution on chromatographic resolution will be evaluated, as well as the effect of temperature on SEC performance. Examples will be given of size-based separations, including separations of monoclonal antibody monomers from aggregates.

INTRODUCTION

METHODSChromatographic Conditions (unless otherwise noted): • Instrument: ACQUITY UPLC® System with UV detection @

280 nm • Temperature: 30˚C • Mobile phase: 100 mM sodium phosphate, pH 6.8 • Flow rate: 0.3 mL/min • Injection type: Full loop, 2 µL • Column dimensions: 4.6x150 mm, except for the Waters

BioSuiteTM 250 4µm UHR column, which was 4.6x300 mm.

Stationary phase particles were synthesized with different total pore volume, mean pore size, surface area, and mean particle size. All materials were diol-bonded to provide a stable chemical surface that exhibited low protein binding.

MATERIALS PROPERTIESFACTORS AFFECTING RESOLUTION IN SEC

The general resolution equation in chromatography is typically defined as Rs≡ΔVR/4σv, where ΔVR is the difference in retention volume between two peaks, and σv is the mean standard deviation of the two peaks in volume units. In SEC, the numerator is representative of the selectivity, while the denominator is a function of the chromatographic efficiency. For a given packing material, a calibration curve can be generated that correlates the retention volume to the molecular weight. A linear relation is obtained for a portion of the curve. We can use the experimentally-determined slope of the curve to quantitate the selectivity of a given chromatographic material. As will be shown, the slope of the calibration curve is directly related to the pore volume of the stationary phase. The chromatographic efficiency is a function of the particle size and packing efficiency.

Table II. Properties of chromatographic materials used in this study.

Analyte pI MW Uracil N/A 112

Ribonuclease A 8.7 13700

Lysozyme 11 14400

Myoglobin 6.8, 7.2 17000

Ovalbumin 4.6 44000

Table I. Test probes used for characterization of SEC columns.

IgG 6.7 150000

Thyroglobulin 4.6 669000

Blue Dextran N/A 2000000

5.8 67000 BSA

EFFECT OF PARTICLE SIZE AND PORE DIAMETER ON EFFICIENCYThe introduction of sub 2 micron particles and Ultra-Performance Liquid Chromatography instrumentation has led to significant improvements in chromatographic efficiency and throughput compared to traditional HPLC separations for reversed-phase chromatography. However, to-date, similar benefits have not been demonstrated for other modes of chromatography, such as size-exclusion chromatography. A recent publication by Held et al (1) states that such benefits cannot be achieved with columns packed with sub-two micron particles for SEC. We evaluated the effect of flow rate on the chromatographic separation of various proteins on columns packed with particles of a range of pore sizes and particle sizes. Retention time and peak width data were first corrected for system effects. Plots of plate height vs. interstitial linear velocity (measured at 50% peak height) are shown in Figures 1 & 2 below. Results show significantly improved performance for the column with 1.7 µm particles compared to 2.6 µm particles. For proteins, the flow rates used were significantly higher than the diffusion-limited region, so data were fit to the linear equation, H=A + C•u. Theory predicts that the A- and C- terms are proportional to 1/dp and 1/dp2, respectively. The experimental data fits well to this expectation. We also found that pore size has limited impact on chromatographic performance as a function of flow rate.

0

20

40

60

80

100

120

0.00 0.20 0.40 0.60 0.80 1.00 1.20 1.40 1.60 1.80 2.00

H (

µm

)

ui (cm/s)

(2) Prototype “D”dp=2.60µm, MPD=191Å

0

20

40

60

80

100

120

0.00 0.20 0.40 0.60 0.80 1.00 1.20 1.40 1.60 1.80 2.00

H (

µm

)

ui (cm/s)

(1) ACQUITY BEH200 SECdp=1.7µm, MPD=211Å

IgG2 dimer IgG2 IgG1 dimer IgG1 Ovalbumin Ribonuclease A Uracil

Figures 1-2. H-u curves for (1) Waters ACQUITY UPLC® BEH200 SEC, 1.7µm, (2) Prototype “D”. Column dimen-sions were 4.6x150mm. Flow rate range 0.1-0.7 mL/min. Other conditions as described in Methods Section.

BEH200 SEC1.7 µm, 203Å

Prototype "D"2.6 µm, 191Å

Prototype "B"2.6 µm, 239Å

Prototype "A"2.5 µm, 297Å

A C r2 A C r2 A C r2 A C r2

Ribonuclease A 3.12 1.36 0.9886 4.33 6.99 0.9982 5.38 6.00 0.9963 5.32 5.73 0.9982

Ovalbumin monomer 15.92 1.74 0.9741 13.30 11.81 0.9991 11.36 25.43 0.9980 10.91 21.99 0.9960

Ovalbumin dimer 30.86 4.23 0.9199 19.19 35.82 0.9949 N.D. N.D. N.D. N.D. N.D. N.D.

IgG1 monomer 10.77 3.83 0.9749 13.93 37.45 0.9973 12.09 9.85 0.9978 10.08 8.88 0.9991

IgG2 monomer 6.80 6.66 0.9953 14.77 43.19 0.9979 19.65 29.17 0.9998 15.16 24.67 0.9982

IgG2 dimer 47.82 13.40 0.9335 24.18 111.91 0.9937 N.D. N.D. N.D. N.D. N.D. N.D.

Table III. A- and C- terms obtained from Figures 1 & 2. Note large A-terms, which may be a result from partial resolution of multiple components in the peak.

BEH200 SEC Prototype "D" Prototype "B" Prototype "A"

A/dp C/dp2 A/dp C/dp

2 A/dp C/dp2 A/dp C/dp

2

Ribonuclease A 2.082 0.607 1.505 0.842 1.873 0.728 1.893 2.038

Ovalbumin monomer 10.611 0.773 4.619 1.424 3.959 3.087 3.882 7.825

Ovalbumin dimer 20.571 1.879 6.663 4.319 N.D. N.D. N.D. N.D.

IgG1 monomer 7.177 1.702 4.836 4.515 4.212 1.196 3.589 3.160

IgG2 monomer 4.534 2.960 5.128 5.207 6.848 3.541 5.395 8.780

IgG2 dimer 31.879 5.955 8.396 13.492 N.D. N.D. N.D. N.D.

Table IV. A/dp and C/dp obtained from Figures 1 & 2.

EFFECT OF PORE VOLUME ON SELECTIVITYRetention times for proteins were determined for columns packed with Materials A-I, which varied in total pore volume from 0.68-1.63 cm3/g, as shown in Table II. A plot of retention volume vs. log MW was plotted for each of the materials (Figure 3). The slope of the each of the curves (in the representative linear region) was determined for each column; Figure 4 shows a plot of the pore volume vs. the slope. As can be seen, a good correlation is obtained between these two factors.

Figure 3. Normalized calibration curves for columns packed with materials as described in Table I. Chromatographic conditions as described in Methods Section. Slopes were obtained from the linear region of each curve.

107

104

105

106

102

103

0.2 0.90.80.70.60.50.40.3

Mole

cula

r W

eight

(Da)

Normalized Retention Volume, VR/V

C

Thyroglobulin

IgG

Dextran Blue

BSAOvalbumin

Cytochrome c

Uracil

Prototype C

BioSuite 250

Prototype I

Prototype H

Prototype G

Prototype F

Prototype D

Figure 4. Plot of normalized slope vs. Specific Pore Volume obtained for each curve shown in Figure 3. Slopes were determined for the linear range of each curve.

Norm

aliz

ed S

lope

from

Cal

ibra

tion C

urv

e, β

Δ(log M

W)/Δ(V

R/V

c)

Pore Volume (cm3/g)

C

I

H

J

G

F

D

The impact of temperature on a chromatographic separation is determined thermodynamically by the expression:

ln k = ΔH˚/RT - ΔS˚/R + ln (φ)where ΔH˚ and ΔS˚ are the respective standard enthalpies and entropies, R is the molar gas constant, T is the absolute tem-perature and φ is the phase ratio of the column. In adsorption chromatography, the entropy term can usually be neglected, as it is negligible relative to the enthalpy term. How-ever, in SEC mode, the opposite is true. Thus in an entropically-driven separation such as SEC, retention should be primarily inde-pendent of temperature. However, chromatographic materials may have ionic or other functional groups on the surface that may induce binding. Ideally, the stationary phase and the chromatographic conditions are opti-mized to prevent adsorption interactions. Examining retention at different temperatures will provide in-formation as to whether adsorption is occurring. Other factors which may contribute to temperature-induced changes in reten-tion, such as changes in conformation or changes in the hydration layer, should also be considered. The chromatograms below show the effect of temperature on the retention of a Protein Test Mix. The top chromatogram was obtained using a BEH-based ACQUITY BEH200 SEC column, while the bottom chromatogram was from a silica-based BioSuite 250 column. Significantly greater retention change is observed for proteins in the latter case. This is attributed to the greater acidity of silanols at the stationary phase surface, which can interact with and adsorb proteins. Note in both cases, uracil retention is affected by temperature. We believe that this is due to H-bond interactions. Further work is being done to better understand this behavior.

INFLUENCE OF TEMPERATURE

REFERENCES1. D Held, G Reinhold and P Kilz, “U-GPC? Making GPC/SEC Faster,” The Column (April 6, 2010) 10-14.

Figure 9. Effect of temperature on the separation of (1) Thyro-globulin, (2) IgG1, (3) BSA, (4) Myoglobin, (5) Uracil. Top: Waters ACQUITY UPLC BEH200 SEC, 1.7 µm, 4.6x150 mm. Bottom: Waters BioSuite 4 µm, 4.6x300 mm.

CONCLUSION We have evaluated columns packed with stationary phases containing a range of mean pore sizes, particle sizes and pore volumes. We have found particle size has a direct impact on chromatographic performance, and correlates well with chromatographic theory. We have also found that the total pore volume correlates well with the slope of the cali-bration curve. By reducing particle size, SEC throughput can be in-creased significantly. In the case of monoclonal monomer/aggregate separations, analysis time can be de-creased dramatically compared to traditional SEC packing materials. Operating temperature was found to have minimal impact on chromatographic retention times for proteins tested on the BEH200 SEC column. On the other hand, retention times were affected by temperature for the BioSuite 250 UHR column, which is silica-based. We attribute this to interac-tions with exposed silanols on the silica surface. Uracil reten-tion is affected by temperature for both columns, which we attribute to hydrogen bonding effects. SEC separations using various buffer types were explored. Salt type was found to affect both retention times and chro-matographic efficiency.

MAb MONOMER/AGGREGATE SEPARATION

The effect of flow rate on the resolution of a monoclonal antibody monomer and dimer was compared for an ACQUITY BEH200 SEC 1.7 µm, 4.6x150mm and a BioSuite 250 UHR 4µm, 4.6x300mm column. Despite the shorter column length of the former column, we found comparable resolution for the two columns when oper-ated at identical flow rates. The smaller particles in the BEH200 provided for even shorter run times by enabling operation at faster flow rates, while still achieving the requisite resolution.

Figures 5 & 6. Effect of flow rate on the separation of a mono-clonal antibody monomer and dimer. Top: Waters ACQUITY UPLC BEH200 SEC, 1.7µm, 4.6x150mm. Bottom: Waters Biosuite 4µm, 4.6x300mm.

Figures 7 & 8. Effect of flow rate on plate height and resolution for separation of monoclonal antibody monomer and dimer. Columns: (1) BEH200 SEC, 1.7µm, 4.6x150, and (2) BioSuite 250 4µm UHR, 4.6x300. Other conditions as described in Methods Section.

BioSuite 250 - dimer

BEH200 - monomer

BEH200 - dimer

BioSuite 250 - monomer

Plat

e hei

ght

(µm

)

Flow rate (mL/min)0.10 0.90.80.70.60.50.40.30.2

0

300

250

200

150

100

50

1.5

2.0

2.5

3.0

0 0.2 0.4 0.6 0.8 1

Res

olu

tion

Flow Rate (mL/min)

BEH200

BioSuite

AU

@ 2

80 n

m

0.00

0.005

Time (minutes)302520151050

BioSuite 250 UHR SEC 4 µm4.6x300mm 0.8 mL/min

0.4 mL/min

0.2 mL/min

0.1 mL/min

0 1412108642 16Time (minutes)

0.00

0.03

0.02

0.01

AU

@ 2

80 n

m

BEH200 SEC 1.7 µm4.6x150mm

0.1 mL/min

0.7 mL/min

0.4 mL/min

0.2 mL/min

AU

0.00

0.04

0.08

0.12

Minutes 2.50 3.00 3.50 4.00 4.50 5.00 5.50 6.00 6.50 7.00

AU

@ 2

80 n

m

0.00

0.04

0.08

0.12

2.50 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5

AU

0.000

0.025

0.050

0.075

0.100

Minutes 5.00 6.00 7.00 8.00 9.00 10.00 11.00 12.00 13.00 14.00

AU

@ 2

80 n

m

0.000

0.025

0.050

0.075

0.100

Time (minutes) 5 6 7 8 9 10 11 12 13 14 15

BEH200 SEC 1.7 µm 4.6x150 mm

BioSuite 250 UHR SEC 4 µm4.6x300 mm

1

5

4

3

2

15

4

32

30°C

40°C

50°C

30°C

40°C

50°C

©2010 Waters Corporation

Material dp (µm) Mean Pore Dia (Å)

Specific Surface Area (m2/g)

Total Pore Vol. (cm3/g)

A 2.5 297 86 0.68

B 2.6 239 108 0.69

C 1.8 139 182 0.69

D 2.6 191 135 0.72

E 1.7 311 217 0.73

F 1.6 348 112 1.12

G 1.6 193 207 1.16

H 2.8 213 225 1.42

I 2.4 259 209 1.63

BEH200 1.7 203 218 1.27

BiosuiteTM 250 4 233 229 1.52

J 2.0 201 217 1.30

INFLUENCE OF SALTS The effect of different salt additives to the mobile phase was explored. Both retention and peak shape were found to be influenced by salt type. Chromatograms are shown below, following the sequence of the Hofmeister Series for anions and cations:

SO42- > PO4

2- > OAc- > Cl- > ClO4-

NH4+ > K+ >Na+

In general, a decrease in retention was observed to follow the Hofmeister series, with the notable exception of sodium chloride. In this case, retention was greater than with other analytes. In addition, peaks were significantly broader.

Figures 10 & 11. Effect of salt type on retention of analytes. Mobile phase consisted of 10 mM sodium phosphate, pH 6.8 with 200 mM of given salt. In the case of phosphate, salt concentration was 100 mM. Analytes same as in Figure 9. Other chromatographic conditions as described in Methods Section.

Abso

rban

ce @

280 n

m (

AU

)

0.00

0.02

0.04

0.00

0.02

0.04

0.00

0.02

0.04

Time (Minutes)

2 3 4 5 6 7 8

Ammonium Chloride

Sodium Chloride

Potassium Chloride

Abso

ban

ce @

280 n

m (

AU

)

0.00

0.02

0.04

0.00

0.02

0.04

0.00

0.02

0.04

0.00

0.02

0.04

0.00

0.02

0.04

Time (minutes) 2 3 4 5 6 7 8

Sodium Sulfate

Sodium Phosphate 100mM

Sodium Acetate

Sodium Chloride

Sodium Perchlorate