BioProcess International™

Analytical and Quality Summit

Application of Light Scattering Techniques for Analysis of

Oligomerization and Particle Formation

Ewa Folta-StogniewYale University

Outline

• Light Scattering Technologies– Static and dynamic light scattering– Parameters derived from SLS and DLS measurements

• Batch Light Scattering Applications– Detection of aggregates in DLS and SLS measurement

• Flow Mode Light Scattering Applications– Molar mass distributions and differences in populations– Characterization of morphology of aggregates

• Determination of an oligomeric state of modified proteins from SEC-LS/UV/RI measurement

• Capabilities and limitation of static and dynamic LS measurements

Light Scattering Experiments

Sample cell

Monochromatic Laser Light

Io I

IΘ

detector

Computer

Θ

Measurements:• batch mode

• “in-line” mode combined with a fractionation step,

i.e. chromatography, mainly Size Exclusion Chromatography, Flow Field Fractionation

• Static (classical)time-averaged intensity of

scattered light

• Dynamic (quasielastic)

fluctuation of intensity of scattered light with time

Light Scattering Experiments

• Static (classical)time-averaged intensity of

scattered light

• Dynamic (quasielastic)

fluctuation of intensity of scattered light with time

Light Scattering Experiments

Parameters derived:• Molar Mass (weight-average) accuracy ~5%

• (<rg2>1/2) root mean square radii

for (<rg2>1/2)> (λ/ 20) ~ 15 nm

• A2 second virial coefficient

Parameters derived:• DT translation diffusion coefficient

• Rh hydrodynamic radius (Stokes radius) Uncertainty of ~10% for monodisperse sample

cAPMR

cKw 22

)(1

)(*

+=θθ

R(Θ) Rayyleigh ratio (excess scattered light)c sample concentration (g/ml)Mw weight-average molecular weight (molar mass)

A2 second virial coefficient (ml-mol/g2)P(Θ) form factor (angular dependence)

2 2 2 4

hT R

kTDπη6

=

Rayleigh-Debye-Zimm formalism Stokes-Einstein

DT translational diffusion coefficientk Boltzmann constantT temperature

Rh radiusη solvent viscosity

because of their big Mw, aggregates scatter strongly even when present at low concentrations; easily detectable

•Scattering Intensity, R(Θ)~ Mw*c

Why Light Scattering?

•Angular variation of the scattered light is related to the size of the molecule

the light scattering signal from aggregates will show angular dependence, while LS signal produces by lower order oligomerslike dimers, trimers, tetramers, et c. will not

•LS measurements are non-invasive and non-destructive•small sample volumes

•great dynamic range for sizing: hydrodynamic radii ~ 2nm to 500 nm

•great dynamic range for Mw determination: < 1kDa to >10 MDa

•wide range of concentrations (non-ideality can be addressed through the determination of second virial coefficient)

•perfectly suited for determination of oligomeric state of modified proteins without prior knowledge of extend of modificaqtion (glycosylated, modified by polyethylene glycol, or membrane proteins present as complexes with lipids and detergents

Peak Rh (nm) Polydispersity (%) MW (R) kDa % Intensity % Mass

1 3.1 12.8 46 54 99.9

2 24 17.8 >1MDa 23 0.1

3 86 13.4 >1MDa 23 <0.1

Ovalbumin; monomer: 43 kDa; Rh=3.0 nm

Rh = 8±7 nm from Cumulant Fit (Polydispersity 93%)

Regularization Fit:

Determination of hydrodynamic radius, Rh, from a Dynamic LS experiment

Peak Rh (nm) Polydispersity (%) MW (R) kDa % Intensity % Mass

1 3.1 12.9 46 96 100

2 32 0 >1MDa 2 0

3 2423 0 >1MDa 2 0

Ovalbumin 43 kDa; Rh=3.0 nm

Rh = 3.2±0.6 nm from Cumulant Fit (Polydispersity 19%)

Regularization Fit:

Results from a batch mode Dynamic LS experiment:

Protein H 23 kDa; Rh=2.3 nm

Dissociation of aggregates upon dilution; time course

Rh=94 nm Rh=23 nm

Rh=2.3 nm; Pd=42%

Rh=2.8 nm

Zimm plot analysis of static light scattering data

Mw = 62 kDaA2 = (5.226 ± 0.316)e-4 mol mL/g²

0.0 0.5 1.0 1.5 2.0-51.60x10

-51.64x10

-51.68x10

-51.72x10

-51.76x10

sin²(theta/2) + 1085*c

K*c

/R(th

eta)

Zimm Plot - BSA_A2_H

RMS : 4.8 ± 1.4 nmMM : (6.180 ± 0.014)e+4 g/molA2 : (5.226 ± 0.316)e-4 mol mL/g²

0.0

2.0

4.0

6.0

8.0

10.0

0.0 2.0 4.0 6.0 8.0

Det

ecto

r: 1

1

Volume (mL)

Baselines - BSA_A2_HK*cR(Θ)

LS signal at 90º

B. H. Zimm, (1948) J. Chem. Phys., 16, 1099

Determination of Molar Mass and second virial coefficient from a batch static LS experiment

BSA 66 kDa

cAPMR

cKw 22

)(1

)(*

+=θθ

0.90

1.00

1.10

1.20

1.30

1.0x10-7 1.0x10-6 1.0x10-5 1.0x10-4 1.0x10-3 1.0x10-2 1.0x10-1 1.0x100 1.0x101

Delay time (sec.)

Temperature : 27.831 °CRh : 3.4 nm

A2

and Rh from DLS

Batch Mode Static MALLS experiment

Sample Weight Average MM, Mw ± SD*[kDa]

RMS[nm]

1 15 ± 1 0

2 126 ± 8 56 ± 10

Monomer 14 kDa

Angular dependence of scattered light clearly indicates presence of aggregates

-1.0

0.0

1.0

2.0

3.0

4.0

5.0

0.0 4.0 8.0 12.0

Det

ecto

r: A

UX

1

Volume (mL)

0.0

0.1

0.2

0.3

0.4

0.5

Det

ecto

r: 1

1

Strip Chart - Sample1_2_aggregates

-56.00x10

-56.50x10

-57.00x10

-57.50x10

-58.00x10

-58.50x10

0.0 0.2 0.4 0.6 0.8 1.0

sin²(Θ/2)

-63.00x10

-64.00x10

-65.00x10

-66.00x10

-67.00x10

0.0 0.2 0.4 0.6 0.8 1.0

sin²(Θ/2)

Sample 1Sample 2

• Static (classical) • Dynamic (quasielastic)

Feature detected in a batch mode LS measurementsfor sample containing aggregates

Aggregates present:

• elevated weight average Molar Mass

(Mw weight average)

• angular dependence in scattered light

Aggregates present:

• autocorrelation function cannot be described by single exponential (cumulant fit)

• polydispersity from cumulant fit >15%

Missing information: how much and what size?

Solutions• Sample fractionation followed by batch measurements

• Column separation with simultaneous LS characterization

sample

HPLC

system

waste or

collection

Computer

UV

detector

LS

detector

DLS+SLS

RI

detector

SEC

column

0.1 μm pre-filtered buffer

0.1 μm “in-line” filter

Three Detector monitoring

-0.5

0.0

0.5

1.0

1.5

2.0

4.0 8.0 12.0 16.0 20.0

LS #

11, A

UX

1, A

UX

2

Volume (mL)

Peak ID - Ova_071305a_01_P_NLS #11AUX1AUX2

UV at 280 nm RI LS at 90°

Ovalbumin 43 kDa

0.2

0.4

0.6

0.8

1.0

0 10 20 30

Det

ecto

r: 1

1

Volume (mL)

-0.5

0.0

0.5

1.0

1.5

2.0

Det

ecto

r: A

UX

1

Strip Chart - OVA_b_UV_traces

88% monomer

8% dimer

1.5% trimer

3% aggregates < 1MDa

0.4% 1-100 MDa

0.4% 1-100 MDa

UV at 280 nm

LS at 90 deg

?

A monomeric protein 43 kDa and aggregates 10 MDa at 2 mg/mL:

Intensity of scattered light ~ Mw*c

due to their high Mw aggregates scatter very strongly

Changes in intensity of scattered light due to aggregates presence

0

1

2

3

0.0% 0.5% 1.0%% (w/w) aggregates

Inte

nsity

of L

S si

gnal

intensity monomer

intentisty aggregates

total intensity

Ovalbumin 43 kDa

3D Plot - OVA_e_RI

34567891011121415161718 163°

90°

14°

Aggregatesangular dependence of scattered light

Lower order oligomersno angular dependence of scattered light

41.0x10

51.0x10

61.0x10

71.0x10

81.0x10

91.0x10

6.0 8.0 10.0 12.0 14.0 16.0

Mol

ar M

ass

(g/m

ol)

Volume (mL)

Molar Mass vs. Volume

OVA_e_UVOVA_200_a_P_N_UV_templat...OVA_b_UVOVA_c_UVOVA_d_UV

Ovalbumin 43 kDa automated template processing of five data sets

Molar mass distribution for multiple analyses

0.0

0.2

0.4

0.6

0.8

1.0

1.0x104 1.0x105 1.0x106 1.0x107 1.0x108 1.0x109

Cum

ulat

ive

Wei

ght F

ract

ion

W

Molar Mass (g/mol)

Cumulative Molar Mass

OVA_e_UVOVA_200_a_P_N_UV_templat...OVA_b_UVOVA_c_UVOVA_d_UV

Determination of Weight Fractions

41.0x10

51.0x10

61.0x10

71.0x10

81.0x10

91.0x10

6.0 8.0 10.0 12.0 14.0 16.0

Mol

ar M

ass

(g/m

ol)

Volume (mL)

Molar Mass vs. Volume

OVA_e_UVOVA_a_UVOVA_b_UVOVA_c_UVOVA_d_UVOva_071305cOva_071305aOva_071305b

Ovalbumin (5 runs)

Mw = 108 ± 17 kDa

Polydispersity Mw/Mn

2.3 ± 0.4

Ovalbumin (3 runs)

Mw = 141 ± 3 kDa

Polydispersity Mw/Mn

2.92 ± 0.06

Differences in population based on molar mass distribution

Differences in population based on molar mass distribution

0.80

0.85

0.90

0.95

1.00

1.0x10 4 1.0x10 5 1.0x10 6 1.0x10 7 1.0x10 8 1.0x10 9

Cum

ulat

ive

Wei

ght F

ract

ion

W

Molar Mass (g/mol)

Cumulative Molar Mass OVA_e_UVOVA_a_UVOVA_b_UVOVA_c_UVOVA_d_UVOva_071305cOva_071305aOva_071305b

Ovalbumin 43 kDa

Ovalbumin (5 runs)

MMw = 108 ± 17 kDa

Polydispersity Mw/Mn

2.3 ± 0.4

Ovalbumin (3 runs)

MMw = 141 ± 3 kDa

Polydispersity Mw/Mn

2.92 ± 0.06

Ovalbumin 43 kDa

Average Mw ± SD

[kDa](5 analyses)

Average Mw ± SD

[kDa](3 analyses)

Fraction of Mass

[% of total](5 analyses)

Fraction of Mass

[% of total](3 analyses)Oligomeric state

Mw = 108 ± 17 Mw = 141 ± 3 Mw = 108 ± 17 Mw = 141 ± 3

Mono (20-50 kDa) 43.0 ± 0.1 42.80 ± 0.02 88.1 ± 0.1 85.23 ± 0.06

Di (50-96 kDa)

Tri (96-130 kDa)Agg. (0.13 –1 MDa)

7.68 ± 0.04

Agg. (1 –100 MDa)

9.4 ± 0.082.7 ± 0.4

1.54 ± 0.05114 ± 4

270 ±10

1.9 ± 0.0

2.87± 0.06

10±1 x103 0.6 ± 0.0

2.18 ± 0.08

0.4 ± 0.0

84.1± 0.2

121.8 ± 0.7

284 ± 2

10.9±0.4 x103

Differences in population based on molar mass distribution

10.0

100.0

1000.0

6.0 7.0 8.0 9.0

R.M

.S. R

adiu

s (n

m)

Volume (mL)

RMS Radius vs. Volume Ova_122105a_Rg-Rh

3D Plot - OVA_e_RI

34567891011121415161718 163°

90°14°

Determination of radius of gyration, Rg, (root mean square radius, R.M.S.,) from angular dependence of scattered light

Morphology of aggregates from angular dependence of LS signal;

size determination- Rg

90° & AUX detector

Peak, Slice : 1, 944 Volume : 7.867 mL Fit degree : 1 Conc. : (1.915 ± 0.020)e-6 g/mL Mw : (2.277 ± 0.024)e+7 g/mol Radius : 46.8 ± 0.2 nm

-84.00x10

-84.50x10

-85.00x10

-85.50x10

-86.00x10

-86.50x10

0.0 0.2 0.4 0.6 0.8 1.0

K*c

/R(th

eta)

sin²(theta/2)

Debye Plot - Ova_b_H_aggregates

Zimm Plot ) )(sin R ) 3/ (16 (11)(

*2

2222g

θλπθ

><+=wMR

cK

Radius: 46.8±0.2 nm

1Rod

0.5Coil

0.33SphereνFor

Rollings, J.E. (1992) in “Laser Light Scattering in Biochemistry”, Eds. S.E. Harding, D. B. Sattelle and V. A. Bloomfield; p. 275-293

Inferring conformational information from the relationship between molecular size (Rg) and molecular weight (Molar Mass)

log(Rg) versus log(MM)

Slope = ν

Rg ~ Mν

10.0

100.0

1000.0

1.0x106 1.0x107 1.0x108 1.0x109

R.M

.S. R

adiu

s (n

m)

Molar Mass (g/mol)

RMS Radius vs. Molar Mass

Ova_122105a_Rg-Rh0.38 ± 0.00

log(

Rg) ν = 0.4

log (MM)

Shape analysis: log(Rg) versus log(MM)

51.0x10

61.0x10

71.0x10

81.0x10

91.0x10

6.0 7.0 8.0 9.0 10.0

Mol

ar M

ass

(g/m

ol)

Volume (mL)

Molar Mass vs. Volume

Ova_Rg-RhAmyloids_peak

10.0

100.0

1000.0

1.0x105 1.0x106 1.0x107 1.0x108 1.0x109

R.M

.S. R

adiu

s (n

m)

Molar Mass (g/mol)

RMS Radius vs. Molar Mass

Amyloids_peak0.75 ± 0.01Ova_Rg-Rh0.38 ± 0.00

ν = 0.4

ν = 0.8

1Rod

0.5Coil

0.33SphereνFor

Amyloids ν = 0.8 Coil/Rod

Ova_aggr ν = 0.4 Sphere/Coil

Aggregates of Ovalbumin vs. “amyloid-type” fibers

For ρ = Rg/Rh

Sphere 0.774

Coil 0.816

Rod 1.732

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

6 7 8 9 10

volume [mL]

0

5

10

15

20

25

30

35

RIRgRh

0

0.2

0.4

0.6

0.8

1

1.2

1.4

6 6.5 7 7.5 8 8.5 9

volume [mL]

0

20

40

60

80

100

120

RIRgRh

Shape analysis: shape factor ρ = Rg/Rh

Aggregates of Ovalbumin vs. amyloid fibers

Shape factor: ρ = Rg/Rh

Combination of MALS (Rg) and DLS (Rh)

Ovalbumin Amyloids

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

6 7 8 9 10

volume [mL]

0

5

10

15

20

25

30

35

RIRgRh

0

0.2

0.4

0.6

0.8

1

1.2

1.4

6 6.5 7 7.5 8 8.5 9

volume [mL]

0

20

40

60

80

100

120

RIRgRh

Shape analysis: shape factor ρ = Rg/Rh

Aggregates of Ovalbumin vs. amyloid fibers

Shape factor: ρ = Rg/Rh

Combination of MALS (Rg) and DLS (Rh)

Ovalbumin Amyloids

B Rizos, A. K., Spandidos, D. A., and Krambovitis, E., (2003) Int. J. Med., 12, 559-563

Rg/Rh = 1.84 Rod

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

6 7 8 9 10

volume [mL]

0

5

10

15

20

25

30

35

RIRgRh

Rg/Rh = 0.91 Coil

0

0.2

0.4

0.6

0.8

1

1.2

1.4

6 6.5 7 7.5 8 8.5 9

volume [mL]

0

20

40

60

80

100

120

RIRgRh

Shape analysis: shape factor ρ = Rg/Rh

Aggregates of Ovalbumin vs. amyloid fibers

Shape factor: ρ = Rg/Rh

Combination of MALS (Rg) and DLS (Rh)

Ovalbumin Amyloids

For ρ = Rg/Rh

Sphere 0.774

Coil 0.816

Rod 1.732

Rg/Rh = 1.84 Rod

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

6 7 8 9 10

volume [mL]

0

5

10

15

20

25

30

35

RIRgRh

Rg/Rh = 0.91 Coil

0

0.2

0.4

0.6

0.8

1

1.2

1.4

6 6.5 7 7.5 8 8.5 9

volume [mL]

0

20

40

60

80

100

120

RIRgRh

Shape analysis: shape factor ρ = Rg/Rh

Aggregates of Ovalbumin vs. amyloid fibers

Shape factor: ρ = Rg/Rh

Combination of MALLS (Rg) and DLS (Rh)

Ova_aggr ν = 0.4 Sphere/Coil Amyloids ν = 0.8 Coil/Rod

AmyloidsOvalbumin

Various uses of Light Scattering for assessing protein aggregates

Experiment Detects Aggregates

Information about population(distribution)

Challenge in use

Sample dilution

Speed

DLS Yes No Low No Fast

Micro-batchMALS

Yes No High No Medium

SEC/MALLS/DLS Yes Yes Medium Yes Medium

Determination of the oligomeric state of modified proteins from SEC-LS/UV/RI analysis

1. Glycosylated proteins

2. Proteins conjugated with polyethylene glycol

3. Membrane protein present as a complex with lipids and detergents

Input:• Polypeptide sequence

• Chemical nature of the modifier

Results:• Oligomeric state of the polypeptide

• Extend of modification (grams of modifier /gram of polypeptide)

“three detector method”

2)(

))((*

RI

UVLSkpMW

ε=

Three Detector Method

Yutaro Hayashi, Hideo Matsui and Toshio Takagi (1989) Methods Enzymol,172:514-28Jie Wen, Tsutomu Arakawa and John S. Philo (1996) Anal Biochem, 240:155-66Ewa Folta-Stogniew (2006) Methods in Molecular Biology: New and Emerging Proteomics Techniques, pp. 97–112

MWp Molecular Weight (polypeptide)

ε extinction coefficient

LS light scattering intensity

UV absorbance (ε)

RI refractive index change

k calibration constant

Protein MW (kDa)

Ova 43BSA(1) 66BSA(2) 132Ald 156Apo-Fer 475

Three-detector calibration 10-17-01

y = 92.383x - 3.4044R2 = 0.9996

0

100

200

300

400

500

0 2 4 6

ratio (LS)*(UV)/(A*(RI^2))

MW

[kD

a]

experimental MWp

Linear (theoretical MWp)

2)(

))((*

RI

UVLSkpMW

ε=

PEG-ylated protein: 75 kDa

36 kDa polypeptide + 39 kDa PEGPolypeptide: 146 kDa

(tetramer: 144 kDa)

Full protein: 291 kDa(tetramer: 300 kDa)

0.0

45.0x10

51.0x10

51.5x10

52.0x10

52.5x10

53.0x10

8.0 10.0 12.0 14.0 16.0 18.0

Mol

ar M

ass

(g/m

ol)

Volume (mL)

Molar Mass vs. Volume

PEG-P_fullPEG-P_pp

PEG-ylated protein: 75 kDa

36 kDa polypeptide + 39 kDa PEG

Polypeptide: 146 kDa(tetramer: 144 kDa)

Full protein: 291 kDa(tetramer: 300 kDa)

0.0

51.0x10

52.0x10

53.0x10

54.0x10

55.0x10

8.0 10.0 12.0 14.0 16.0 18.0

Mol

ar M

ass

(g/m

ol)

Volume (mL)

Molar Mass vs. Volume

BSAAPOPEG-P_fullPEG-P_pp

475 66 kDa

Dynamic LS• very fast detection of aggregates• great dynamic range• well suited to study kinetics of aggregation • DLS detector available in a plate reader format for high volume analyses

CapabilitiesStatic LS• fast and accurate determination of molar masses (weight average)

– glycosylated protein, conjugated with PEG, protein-lipids-detergent complexes, protein-nucleic acid complexes

• accuracy of ± 5% in Molar Mass determination• easy to implement, fully automated (data collection and data analysis) • highly reproducible (no operator bias)• SEC/MALS excellent in detecting and quantifying population with various oligomeric state

in protein

Combined data about MM, Rg and Rh - shape information (multiangle static and dynamic LS)

• via frictional ratio Rh/Rs

• via shape factor ν, from log(Rg) vs. log(MM) plot

• via shape factor ρ, from Rg/Rh ratio

Static LS• measures weight average molar mass – needs fractionation to resolve

different oligomeric states • possible losses of sample during filtration and fractionation• limitation on solvent choices (related to a fractionation step)• SEC/SLS/DLS dilution during experiment

Dynamic LS• measures hydrodynamic radius, which is affected by shape

• cannot discriminate between shape effects and changes in oligomeric states, i.e. non-spherical shape mimics oligomerization

• needs fractionation to resolve low number oligomers when present in mixture

Limitations

Ken WilliamsDirector of W.M. Keck Biotechnology Resource

Laboratory at Yale University School of Medicine

NIH

Users of SEC/LS Service

http://info.med.yale.edu/wmkeck/biophysics

Ewa.Folta-Stogniew@yale.edu