Analysis of Protein Aggregation, Stability, and Lot Comparability by Sedimentation Velocity

gayoldΜηχανική

21 Φεβ 2014 (πριν από 3 χρόνια και 7 μήνες)

115 εμφανίσεις

Analysis of Protein Aggregation, Stability, and Lot Comparability by Sedimentation Velocity John S. Philo, Yoshiko Kita and Tsutomu Arakawa Alliance Protein Laboratories, 3957 Corte Cancion, Thousand Oaks, CA 91360 805-388-1074, fax 805-388-7252, http://www.ap-lab.com
ABSTRACT:
Sedimentation velocity is a powerful tool for detecting and quantifying aggregates in protein
pharmaceuticals and proving comparability of conformation and aggregate content for different manufacturing processes or lots. Recent software techniques, including new approaches developed in our lab, extend the power of the method and make the data easier to interpret.Several examples based on studies of antibodies for our clients will be presented. Comparing several lots of monoclonals from two different manufacturing processes demonstrated that the conformation of the majority species is indistinguishable, but different lots contained from 2.0% to 6.0% of dimers and higher aggregates. This aggregate content is significantly higher than detected by SEC, due to aggregate loss on the column matrix. One significant advantage of sedimentation velocity is the ability to run samples under a wide range of solvent conditions. By testing another antibody directly in two different formulation solvents we showed that a low ionic strength formulation gives markedly less dimer, a difference that could not be detected under the high ionic strength conditions necessary for SEC.A common problem in protein products is the irreversible accumulation of aggregates with time or thermal stress. When these aggregates precipitate or form very large particles (“snow”) they are easy to detect, but smaller non-covalent soluble aggregates (which are often precursors to “snow”) can be difficult to detect and quantitate. Using sedimentation velocity on heat stressed samples we studied variations in aggregate amount and size distribution with various additives. These studies also detected a significant change in antibody conformation under conditions that minimize aggregate formation, but it is unclear whether this conformation change is directly a cause of improved reversibility of unfolding.
Objectives
1.Use sedimentation velocity to characterize the size and
mass fraction of aggregates in bulk manufacturing or in-
process samples to:
a.detect lot-to-lot variations
b.help identify steps where aggregates are formed
c.demonstrate equivalence between different processes
2.Characterize aggregates in protein samples under
formulation conditions and after thermal stress to help
optimize the formulation
3.Measure aggregation induced by physical stresses during
manufacturing or storage such as shear-or surface-
induced denaturation
4.Develop improved data analysis methods to support these
goals
Methods
Sedimentation velocity studies were done in a Beckman
Optima XL-A analytical ultracentrifuge using 2-channel
charcoal epon centerpieces. For most of the studies here the
rotor speed was 45,000 rpm, protein concentration was ~0.5
mg/ml, and absorbance scans were done at 280 nm.
Data analysis was done using the program DCDT+ written by
John Philo.
1
1Philo, J.S. (2000). A method for directly fitting the time derivative of
sedimentation velocity data and an alternative algorithm for calculating
sedimentation coefficient distribution functions.
Anal. Biochem.
, in press.
6.06.16.26.36.46.56.66.76.86.97.07.1
Radius (cm)
-0.1
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
1.1
1.2
Absorbance
Some raw velocity data for an antibody---how can we interpret it?
the “dc/dt method” uses a
group of closely-spaced
scans which are subtracted
in pairs to approximate the
time derivative of the data
and thereby derive how
much material is
sedimenting at various rates
The rate of motion of the sedimentation boundary as the run proceeds is determined
by the sedimentation coefficient, which depends on both mass andconformation. If
the sample contains irreversible aggregates, they will form separate boundaries.
a second
boundary?
2468101214
0.0
0.1
0.2
0.3
0.4

g(s*)
, AU/Svedberg
s*, Svedbergs
The sedimentation coefficient distribution function for this antibody
sample shows it is heterogeneous and contains at least 3 species
24681012
0.0
0.1
0.2
0.3
the ratio
s/D
gives the mass
peak width
gives diffusion
coefficient,
D
peak center gives
sedimentation
coefficient,
s
data
fit as one
species

g(s*)
, AU/Svedberg
s*, Svedbergs
Further interpretation is possible by fitting individual speciesas Gaussians;
this is an example for a homogeneous antibody sample
Results

Sedimentation velocity was used to study comparability
between different manufacturing processes for a monoclonal
antibody as well as lot-to-lot variability
–the results showed previously undetected lot-to-lot variability
in aggregate content, but the difference between processes
was no larger than variations within each process
–since sedimentation coefficients are very sensitive to subtle
changes in conformation, the fact that the main peaks from
both processes have the same sedimentation coefficient
proves equivalence of solution conformation

Note that analysis of these samples by size-exclusion
chromatography (SEC) underestimates the true aggregate
content
–this appears to be due to the aggregates sticking to the SEC
column
Comparability testing of two manufacturing processes for a
monoclonal antibody: the main peaks have identical
conformation, but aggregate levels vary lot-to-lot
24681012
0.0
0.1
0.2
0.3
lot B is 94.0% main peak
(96.7% by SEC)


g(s*) (AU/Svedberg)
sedimentation coefficient (Svedbergs)
24681012
0.0
0.1
0.2
0.3
0.4
0.5
data
single species fit
lot A is 98.0% main peak
(99.0% by SEC)



Results (continued)

Comparison of different formulations of an antibody showed
that one contains significantly less dimer
–The low ionic strength of the better formulation would not
allow SEC studies under those formulation conditions

A new software method developed in our lab indicates there
are actually two types of antibody dimer in these samples:
–a slowly-sedimenting extended conformation (probably end-
to-end)
–a rapidly-sedimenting compact conformation (probably side-
by-side)
Velocity analysis of two different formulations of an
antibody, each analyzed in its own formulation buffer,
reveals differences in aggregation
24681012
0.0
0.1
0.2
0.3
0.4
0.5
a low ionic strength
formulation produces
significantly less dimer
shift of main peak is due to
differences in buffer viscosity


g(s*)
(AU/Svedberg)
sedimentation coefficient (Svedbergs)
Detailed analysis of the heterogeneous antibody sample reveals 3species.
Because light scattering data shows that only monomer and dimer are
present, we believe there are two different conformations of dimer
2468101214
0.0
0.1
0.2
0.3
0.4
data
species 1 (monomer)
species 2 (dimer)
species 3 (dimer)
sum

g(s*)
, AU/Svedberg
s*, Svedbergs
Results (continued)
•Studies of monoclonal samples after thermal stress
reveal strong differences among buffers in formation of
soluble aggregates
–The glycine formulation which gives the least aggregation
also produces a significantly different sedimentation
coefficient for the main peak (different conformation);
whether this conformational difference is directly related to
the improved stability is not clear.
•Sedimentation velocity also detected formation of
soluble aggregates after filtration (presumably induced
by shear forces or exposure to the large filter surface)
Formulation applications: analysis of antibody samples after
accelerated stability studies in various buffers
0510152025303540
0.0
0.1
0.2
0.3
0.4

19 mM glycine, pH 6.0
19 mM citrate, pH 6.5
19 mM Tris, pH 7.0
19 mM histidine, pH 6.5
19 mM phosphate, pH 6.5
g(s*) [AU/Svedberg]
s* [Svedbergs]
Assessing effects of physical stress: the homogenous
antibody sample shows aggregates after filtration
4681012
0.0
0.1
0.2
0.3
0.4
data
fit as 1 species
aggregates created
by filtration

g(s*)
, AU/Svedberg
s*, Svedbergs
Conclusions
•Sedimentation velocity is a very useful method for
detecting protein aggregates (both covalent and non-
covalent), characterizing their size, and quantitating the
aggregate content
•Sedimentation velocity is also very useful for
demonstrating comparability of solution conformation for
the main species
•A major advantage of this method over size-exclusion
chromatography is that the analysis can usually be done in
the actual formulation buffer, and there is no concern about
loss of aggregates to a column matrix
•The utility of this technique has grown considerably with
modern instrumentation and improved software tools