Hos 2014 Presentation, February 2014, John Philo

1. Studying Reversible Self-Association of
Biopharmaceuticals using AUC and Light
Scattering
John Philo
HOS2014, February 2014
© Copyright 2014 Alliance Protein Laboratories

2. Outline
 Some examples of applying sedimentation
equilibrium (SE) and sedimentation velocity (SV) to
study reversible self-association
 for each, how do we distinguish reversible from
irreversible association/aggregation?
 Using classical light scattering to study reversible
interactions (CG-MALS)
 the complex world of “crowded solutions” at
~100 mg/mL

3. 6.30 6.35 6.40 6.45 6.50 6.55 6.60
0.0
0.2
0.4
0.6
0.8
1.0
cell base
Absorbance
radius (cm)
diffusion
buoyancysedimentation
meniscus
The fundamentals of sedimentation equilibrium
The concentration
distribution depends
only on molecular
weight, independent
of shape!
 smaller sample size to reduce time to reach equilibrium 

4. Apparent mass vs. concentration plots for the soluble extracellular
domain (ECD) of the atrial natriuretic peptide receptor (monomer
mass 58 kDa)
0.004 0.01 0.02 0.03 0.04 0.1 0.2
0
20000
40000
60000
80000
100000
120000
Mw
concentration (mg/ml)

5. Global analysis of all the data shows the ECD dimerizes with
Kd = 520 +/- 20 nM (G = -8570 +/- 25 cal/mol)
-0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8
0.0
0.3
0.6
0.9
Absorbance(280or230nm)
(r
2
- r0
2
) / 2

6. Reversible or not? This monomer-dimer-tetramer system acts as
a mixture at low concentrations (? due to dilution of its cofactor)
0.005 0.010 0.020 0.050 0.100 0.200 0.500
Concentration (mg/ml)
-4.0E+4
-2.0E+4
0.0E+0
2.0E+4
4.0E+4
6.0E+4
8.0E+4
1.0E+5
1.2E+5
1.4E+5
1.6E+5
1.8E+5
2.0E+5
2.2E+5
2.4E+5
2.6E+5
ApparentMass(Da)
00004.RA2 Ch A-BL 00004.RA2 Ch B 00004.RA2 Ch C 00007.RA2 Ch A-BL
00007.RA2 Ch B 00007.RA2 Ch C 00004.RA3 Ch A 00004.RA3 Ch B
00004.RA3 Ch C 00007.RA3 Ch A 00007.RA3 Ch B 00007.RA3 Ch C

7. SE data for a peptide (3-5 kDa) which strongly reversibly
self-associates; above ~3 mg/mL repulsive solution non-
ideality causes the apparent mass to decrease
0.3 1 10 20
0
1
2
3
4
5
Mw
(#ofsubunits)
concentration (mg/mL)

8. Sedimentation velocity of reversibly-
associating proteins/peptides

9. First we must test for reversible association by doing a dilution
series; here is such a test for a vaccine drug substance that
contains a wide range of different oligomers (up to ~25-mer)
0 5 10 15 20 25 30 35
0.00
0.02
0.04
0.06
0.08
0.10
1.2 mg/mL, sw
= 10.3 S
0.6 mg/mL, sw
= 10.1 S
0.3 mg/mL, sw
= 10.3 S
normalizedg(s*)[perSvedberg]
sedimentation coefficient [Svedbergs]

10. Applying separation methods to reversibly-
associating proteins raises issues which don’t arise
for measurements at equilibrium
 Separation methods (SEC, sedimentation velocity, FFFF,
electrophoresis) typically will not resolve individual oligomer
species for a molecule in rapidly-reversible association
equilibrium
 what is detected will depend on the rates of association and
dissociation compared to the speed of separation
 for associations to oligomers larger than dimer, from theory
multiple peaks may be seen even for infinitely-fast kinetics,
but those peaks usually do not represent single oligomers
 theory says it is possible to see more peaks than the number
of different species that are present!
 The bottom line: your so-called “dimer” or “trimer” peak may
really represent a dynamic mixture of different rapidly-reversible
oligomers

11. Here is a SV dilution series for a monoclonal antibody
that self-associates strongly in a high ionic strength SEC
mobile phase
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
0.0
0.5
1.0
0.2 mg/mL; sw
= 7.61 S
0.5 mg/mL; sw
= 7.99 S
2.0 mg/mL; sw
= 9.44 S
normalizedc(s)[perSvedberg]
sedimentation coefficient, s20,w
[Svedbergs]

12. Studying reversible interactions using
classical light scattering (CG-MALS)

13. Concentration Gradient MALS combines computerized syringes and
a mixing system that flows the mixed solutions into a standard
MALS detector and then on into a concentration detector
concentration
detector (RI or UV)

14. This is a 10-step dilution from ~1.5 mg/mL of that
strongly associating mAb in its SEC mobile phase
1
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
0.73
0
0.1
0.2
0.3
0.4
0.5
0.6
Slice #
45000.000 500.0 1000 1500 2000 2500 3000 3500 4000
Det. 2
Det. 3
Plateaus
UV297
AnalysisSignals

15. CG-MALS results from 3 such dilution series show that the weight-
average molar mass is actually above that of dimer at 1.5 mg/mL
and falls continuously as the concentration falls
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6
0
50
100
150
200
250
300
350
400
apparentmolarmass(kDa)
total protein concentration (mg/mL)
combined data from three 10-step dilution experiments

16. The CG-MALS data fit fairly well to a monomer-dimer-
tetramer equilibrium
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6
0.0
0.2
0.4
0.6
0.8
monomer
dimer
tetramer
speciesconcentration(mg/mL)
total protein concentration (mg/mL)
individual species concentrations from the best-fit model
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6
0
2
4
6
8
10
12
14
16
best fit of mAb-X CG-MALS data
LSsignal(arb.units)
total protein concentration (mg/mL)
experiment
monomer  dimer  tetramer fit

17. The next frontier---weak associations at
very high concentrations (50-200 mg/mL)

18. Light scattering of monomeric HSA up to ~110 mg/mL
(physiological pH, ionic strength) shows very strong
non-ideality
18
~10 mg/mL
~110 mg/mL
Semi-dilute
slide courtesy Mike Marlowe
2
2 3
app true
1 1
...B c B c
M M
   

19. Monomeric fit with non-ideality is clearly insufficient;
indicates presence of higher order species
19
Effective Hard Sphere Model
Virial Coefficient Model
slide courtesy Mike Marlowe

20. Weak self-association of HSA monomers into dimers
starts around 30 mg/mL (~0.5 mM), and then higher
oligomers also form
20
90° LS
Dimer
Tetramer
Monomer
slide courtesy Mike Marlowe

21. Some remaining challenges
1. None of the analytical methods are good at handling
samples containing both reversible and irreversible
oligomers (which is often the real situation)
2. For these samples at very high concentrations can we
really quantitatively account for the non-ideality well
enough to reliably measure weak reversible
interactions? How can we cross-check such results to
prove they are right?
3. CG-MALS requires ~5-10 mL of the solution at high
concentration, which is likely prohibitive except for
late stage development
 can sedimentation equilibrium get us the same information
with much less material?