[PPT] - Water NMR for Protein Aggregation Characterization Marc Taraban PowerPoint Presentation

SLIDE 1

Water NMR for Protein Aggregation Characterization

Marc Taraban

Department of Pharmaceutical Sciences University of Maryland School of Pharmacy

December 5, 2016, Baltimore Protein Aggregation Measurement in Biotherapeutics: Established and Emerging Techniques

SLIDE 2

Water NMR—A Nuisance or A Tool?

BSA (15 mg/mL in PBS buffer) with water suppression (100 scans): bad resolution

f

protein resonances. BSA (15 mg/mL in PBS buffer) without water suppression (1 scan): high S/N for narrow water signal, protein resonances are invisible.

Y. Feng, M. Taraban & Y.B. Yu (2015) Chem. Commun. 51, 6804

In aqueous solutions, solute resonances are considered the most important for NMR, and deuteration or suppression is used to remove interfering water signal

SLIDE 3

Relaxation Rates in NMR

http://www.chem.ch.huji.ac.il

Relaxation

f

the nucleus to its ground state (aligned with external magnetic field) is controlled by two mechanisms.

T 1 T 2

T1 is spin-lattice or longitudinal relaxation, equal to time of

energy transfer from excited to ground state along z-axis,

ften is defined by interaction between nucleus and media

(solvent, diffusion).

T2 is spin-spin or transverse relaxation, equal to time of

energy transfer within the nucleus in the xy-plane due to dephasing, NMR line broadening down to the disappearance

f the NMR signal, often defined by dipolar interactions,

anisotropy of molecule, etc.

SLIDE 4

Transv ansver erse Relax se Relaxation of W tion of Water CPMG Pulse CPMG Pulse Seq Sequence uence

T2 is measured using classic CPMG pulse sequence that allows to monitor the drop in

magnetization in xy-plane.

R2(1H2O) = 1/T 2

SLIDE 5

Water NMR—A Tool

Y. Feng, M. Taraban & Y.B. Yu (2014) Chem. Commun. 50, 12120

Shear Modulus G, kPa

R2(1H2O), s-1

Water signal carries information on the global changes in the solute—water molecules interact with solute molecules and become sensitive to its changes, e.g., association Water proton transverse relaxation rate, R2(1H2O), could be used to measure the stiffness of peptide-based hydrogel. Gelation and aggregation both involve association, so would R2(1H2O) also be sensitive to protein aggregation?

SLIDE 6

Daszkiewicz et al. (1963) Nature, 200, 1006; Oakes et al. (1976) J. Chem. Soc. Faraday Trans. I 72, 228; Hills et al. (1989) Mol. Phys. 67, 903; Indrawati et al. (2007) J. Sci. Agric. 87, 2207

20 40 60 80 100 2 4 6 8 10

R2(

1H2O), s

1

C, mg/mL

20 40 60 80 100 0.4 0.8 1.2 1.6

R1(

1H2O), s

1

C, mg/mL

0.007 s-1.(mg/mL)-1 0.106 s-1.(mg/mL)-1 6

Ovalbumin  fresh;  heat denatured (0.33 Tesla, 14 MHz 1H) Does R2(1H2O) correlate with aggregate size?

Prior Art—Water Relaxation in Heat-Denatured Proteins

0.004 s-1.(mg/mL)-1 0.014 s-1.(mg/mL)-1

SLIDE 7

Y. Feng, M. Taraban & Y.B. Yu (2015) Chem. Commun. 51, 6804

A Probe for Protein Aggregation

Bovine Serum Albumin γ-Globulin

R2(1H2O) linearly increases with the growth of average hydrodynamic radius of

protein aggregates. Similar sensitivity observed in high (400 MHz) and low-field (20 MHz, BT NMR) Water proton NMR is sensitive towards heat-induced aggregation of BSA and human γ-globulin, and could be used to quantify protein aggregation

4 6 8 10 12 0.6 0.8 1.0 1.2 400 MHz 20 MHz 10 20 30 40 0.6 0.8 1.0 1.2 400 MHz 20 MHz

Rh, nm R2(1H2O), s-1

SLIDE 8

mAb has been stressed by: Freeze-Thaw (-40°C  5°C, 16 cycles) Heating at 50°C (36 h) Agitation (24 h) Aggregation was studied by Conventional Techniques Size-Exclusion Chromatography (SEC) Dynamic Light Scattering (DLS) Micro-Flow Imaging (MFI) & Water NMR (wNMR) Transverse Relaxation Rate of Water

R2(1H2O)

Generation of Monoclonal Antibody Aggregates of various sizes

Aggregates 0.45 to 5 µm Aggregates ≥ 5 µm Aggregates ≤ 0.45 µm

SLIDE 9

Measurement of mAb Aggregation by wNMR

R2(1H2O) responded to aggregate formation under different stresses and differs

from control after filtration

R2(1H2O) increased in each stressed sample compared to the unstressed

control sample

Filtration reduced the increase in R2(1H2O) for all stresses
R2(1H2O) was still different after 0.45 micron filtration between stresses

1.05 1.10 1.15 1.20 1.25

R2(

1H2O), s

1

C

n

t r

l

N

n

f i l t e r e d 5

 

m f i l t e r e d . 4 5

 

m f i l t e r e d C

n

t r

l

N

n

f i l t e r e d 5

 

m f i l t e r e d . 4 5

 

m f i l t e r e d C

n

t r

l

N

n

f i l t e r e d 5

 

m f i l t e r e d . 4 5

 

m f i l t e r e d

Freeze-Thaw Heating Agitation

SLIDE 10

Measurement of mAb Aggregation by SEC

4 6 8 10 12 14 200 400 600 800 1000 Control

mAU Time, min

4 6 8 10 12 14 200 400 600 800 1000

mAU Time (min)

Freeze-Thaw Heating, 50°C Agitation

4 5 6 7 8 50 100 150

mAU Time (min)

Freeze-Thaw Heating, 50°C Agitation

Control Stressed samples Aggregates

Aggregates Control Nonfiltered 5 m filtered 0.45 m filtered Freeze- Thaw % LMW 0.8 7.4 7.3 7.3 % HMW 0.0 1.9 2.0 2.0

Total % Aggr 0.8 9.3 9.3 9.3

Heating 50°C % LMW 0.8 1.4 1.4 1.4 % HMW 0.0 6.5 6.4 6.4

Total % Aggr 0.8 7.9 7.8 7.8

Agitation % LMW 0.6 3.2 3.2 3.4 % HMW 0.0 6.9 6.0 5.9

Total % Aggr 0.6 10.1 9.2 9.3

% Low (LMW) and % High Molecular Weight (HMW) and Total % soluble mAb aggregates for three stresses

Total percentage of aggregates were similar, but aggregate

profile was different between each stress type

5 µm & 0.45 µm filtration did not change the ratio between

LMW and HMW aggregates or total percentage of aggregates

SLIDE 11

Measurement of mAb Aggregation by MFI

10

1

10

2

10

3

10

4

10

5

10

6

 25 m  10 m  5 m  2 m Control

Particle Concentration (#/mL) Nonfiltered 5 m filtered 0.45 m filtered

 1 m  1 m  2 m  5 m  10 m  25 m  1 m  2 m  5 m  10 m  25 m  1 m  2 m  5 m  10 m  25 m

10

1

10

2

10

3

10

4

10

5

10

6

Particle Concentration (#/mL) Control Nonfiltered 5 m filtered 0.45 m filtered

 1 m  2 m  5 m  10 m  25 m  1 m  2 m  5 m  10 m  25 m  1 m  2 m  5 m  10 m  25 m  1 m  2 m  10 m  5 m  25 m

10

1

10

2

10

3

10

4

10

5

10

6

Particle Concentration (#/mL) Control Nonfiltered 5 m filtered 0.45 m filtered

 1 m  2 m  5 m  10 m  25 m  1 m  2 m  5 m  10 m  25 m  1 m  2 m  5 m  10 m  25 m  1 m  2 m  5 m  10 m  25 m

5 µm filtration reduced particle counts from ≥ 1 µm to ≥ 25 µm (not only for ≥ 5 µm

particles)

After 0.45 µm filter, the samples for all three stresses are very close to the unstressed

control

Freeze-Thaw Heating, 50 °C Agitation Decrease in particle counts during filtration from MFI

SLIDE 12

Measurement of mAb Aggregation by DLS

5 µm and 0.45 µm filters had minimal effect on PSD for freeze-thaw and heating
Both filtration steps affected the PSD for agitation stress significantly

Each stress produced different particle size distributions (PSD) of aggregates

Freeze-Thaw Heating 50°C Agitation

0.45 µm filtered 5 µm filtered Nonfiltered

0.00 0.02 0.04 0.06 0.05 Intensity, arb. unit 10 10

1

10

2

10

3

10

4

0.0 0.1 0.2 Rh, nm 0.05 0.10 0.05 0.10 0.15 Intensity, arb. unit 10 10

1

10

2

10

3

10

4

0.00 0.05 0.10 Rh, nm

0.0 0.1 0.2 0.3 0.2 0.4

Intensity, arb. unit 10 10

1

10

2

10

3

10

4

0.0 0.2 0.4 Rh, nm

10 10 1 10 2 10 3 10 4 0.0 0.2 0.4 10 10 1 10 2 10 3 10 4 0.0 0.2 0.4 10 10 1 10 2 10 3 10 4 0.0 0.2 0.4

SLIDE 13

Sensitivity of Each Technique to Size and Number of Aggregates

0.0 0.2 0.4 0.6 0.8 1.0

Normalized Soluble Aggregates (arb. unit)

N

n

f i l t e r e d 5

 

m , f i l t e r e d . 4 5

 

m , f i l t e r e d N

n

f i l t e r e d 5

 

m , f i l t e r e d . 4 5

 

m , f i l t e r e d N

n

f i l t e r e d 5

 

m , f i l t e r e d . 4 5

 

m , f i l t e r e d

Freeze-Thaw Heating Agitation

0.0 0.2 0.4 0.6 0.8 1.0

Normalized R2(

1H2O) (arb. unit)

N

n

f i l t e r e d 5

 

m , f i l t e r e d . 4 5

 

m , f i l t e r e d N

n

f i l t e r e d 5

 

m , f i l t e r e d . 4 5

 

m , f i l t e r e d N

n

f i l t e r e d 5

 

m , f i l t e r e d . 4 5

 

m , f i l t e r e d

Freeze-Thaw Heating Agitation

0.0 0.2 0.4 0.6 0.8 1.0

Normalized Particle Concentration,  1m, arb. unit

Freeze-Thaw Heating Agitation

N

n

f i l t e r e d 5

 

m , f i l t e r e d . 4 5

 

m , f i l t e r e d N

n

f i l t e r e d 5

 

m , f i l t e r e d . 4 5

 

m , f i l t e r e d N

n

f i l t e r e d 5

 

m , f i l t e r e d . 4 5

 

m , f i l t e r e d

SEC detects only soluble aggregates: Minimal sensitivity to sample filtration Removal of larger particles seen by MFI 5 µm filtration had major impact

n R2(1H2O), lesser change after

0.45 µm filtration, but differences still seen after 0.45 um filtration

Data are normalized by the difference between stressed sample and control, so as fully stressed sample corresponds to 1, and control corresponds to 0.

SEC SEC wNMR wNMR MFI MFI

SLIDE 14

Sensitivity range of each method to antibody aggregates

wNMR was most consistently sensitive to differences in sample quality across each stress type and after sample filtrations

SLIDE 15

Possible Mechanisms of Sensitivity of Proton NMR to Protein Aggregates

SLIDE 16

Water NMR Senses N r NMR Senses Nanoparticle Clus noparticle Clustering ring

100 1000 10000 0.50 0.55 0.60 0.65 0.70

R2(

1H2O), s

1

Diameter, nm

But one of them demonstrate anomalously high water relaxation rate

WHY? WHY?

Two 200 nm polystyrene nanoparticle samples are visually indistinguishable
1H NMR spectra show no difference in the signal intensities or chemical shifts

SLIDE 17

1000 2000 3000 4000 5000 0.0 0.2 0.4 0.6 0.8 1.0

10

4

10

3

10

2

10

1

10

1

10

1

10

3

10

5

EXPERIMENT DATA FIT

I (cm

1sr
1)

q (Å

1)

Volume Distrubution, arb. unit Diameter, nm

100 200 300 400 500 600 700 800 0.0 0.2 0.4 0.6 0.8 1.0

10

4

10

3

10

2

10

1

10

1

10

1

10

3

10

5

EXPERIMENT DATA FIT

I (cm

1sr
1)

q (Å

1)

Volume Distrubution, arb. unit Diameter, nm

The anomalous sample had quality issues (confirmed by manufacturer), and nanoparticles in this sample are clustered and formed larger assemblies

Water NMR Senses N r NMR Senses Nanoparticle Clus noparticle Clustering ring

USAXS shows that anomalous sample contain mainly 2-4 µm particulates While the good quality sample overwhelmingly contains 200 nm particles

USAXS = Ultra-small angle X-ray Scattering

SLIDE 18

100 1000 10000 0.50 0.55 0.60 0.65 0.70

R2(

1H2O), s

1

Diameter, nm

Sample polydispersity is not affecting R2(1H2O)— mixture

f

different sizes is close to monodisperse sample/blank buffer (blue line).

Water NMR Senses N r NMR Senses Nanoparticle Clus noparticle Clustering ring

= <<

Water molecules in the clustered compartments have different diffusive exchange and local magnetic field gradient resulting in anomalously high

R2(1H2O)

SLIDE 19

Water transverse relaxation rate R2(1H2O) was a

sensitive probe responding to changes in solute molecules: association, clustering, aggregation, etc.

In protein aggregation, R2(1H2O) was sensitive to

the presence of insoluble particulates ≥ 5 µm, from ≥ 1 µm to 5 µm as well as to soluble protein aggregates below 1 um.

R2(1H2O) can be monitored noninvasively using

inexpensive benchtop low-field NMR spectrometers with wide bore capable to accommodate drug product vials without opening or sampling.

Conclusions Conclusions

DON'T THROW THE BABY OUT WITH THE BATHWATER!

SLIDE 20

Acknowledgements

UMB School of Pharmacy Rene Cosme Soh Fongang (PharmD) Huy Chang Truong (PharmD)

Dr. Yue Feng
Prof. Y

. Bruce Yu MedImmune

Dr. Brian Lobo (Formulation)
Dr. Roberto A. DePaz (Formulation)

Light Scattering Center University of Maryland Supported by through