|
Peter J. Morgan, Olwyn D. Byron, Stephen E. Harding
Department of Applied Biochemistry and Food Science University of Nottingham Sutton
Bonington, U. K.
Introduction
One of the latest attempts to improve monoclonal antibodies in their use
as radiolabeled diagnostic markers(1) has involved
the cross-linking of novel Fab' fragments. It is expected that a relationship exists
between the spacing of the antigen binding sites at the extremes of the Fab' fragments in
solution and their immunological performance in vitro and in vivo.
In order to properly understand the function of these new antibody
fragments in vivo, it is critical to have an appreciation of their size,
self-association behavior (or preferably lack of) and solution conformation. Appropriate
application of the Optima XL-A analytical ultracentrifuge can provide this data. No other
single technique provides this breadth of information.
The flexibility conferred upon the cross-linked F(ab')2 by
its linker precludes the determination of its conformation via protein x-ray
crystallography. However, the sedimentation coefficients of both the F(ab')2
and Fab' fragments when combined with radius of gyration data obtained from small angle
x-ray scattering experiments provide a useful gauge of solution conformation.
In this study the monodispersity and absence of self-association
phenomena of Fab' and F(ab')2 solutions are strongly indicated by sedimentation
velocity and sedimentation equilibrium experiments. The weight average molecular weights
measured are shown to be in complete agreement with the molecular weights as calculated
from amino acid sequences. Finally, sedimentation coefficients have been measured, and we
indicate how these can be used to access conformation when combined with other solution
measurements.
Methods
The Fab' and (Fab')2 fragments (kindly supplied by Celltech
Ltd.) were dissolved in a standard phosphate chloride buffer (I = 0.1, pH = 6.8). The
Beckman Coulter Optima XL-A analytical ultracentrifuge was employed to perform both
sedimentation velocity and sedimentation equilibrium experiments. Solute distributions at
20.0°C were recorded via their absorption at 278 nm.
Sedimentation Velocity
Consecutive scans were recorded at regular intervals, utilizing the
"autoscan" facility. Sedimentation coefficient (s20,w) values
were determined in the standard way and plotted against concentration (corrected for
radial dilution). Extrapolation to infinite dilution yielded s20,w.
Sedimentation Equilibrium
The low-speed sedimentation equilibrium method was employed. The
low-speed method is more accurate since it provides more data points over the entire
radial pathlength compared with the high-speed sedimentation equilibrium method (see e.g.,
Ref. 2.) It was considered that equilibrium had been established
when two consecutive scans, recorded several hours apart, appeared identical. Multichannel
(Yphantis type(3)) centerpieces were employed to
permit the simultaneous measurement of multiple samples. The final solute distribution
ASCII data were captured and analyzed on the IBM 3084Q Phoenix mainframe at the University
of Cambridge, using the FORTRAN MSTARA program.(4)
Whole-cell weight average molecular weights (M0w) were
extracted by using the limiting value at the cell base of the M* (point average molecular
weight) function.(2) (An independent estimate for the
initial loading concentration was not required.) Partial specific volumes were calculated
from the amino acid sequence for the fragments (see e.g., Ref.
5).
Results and Discussion
Confirmation of Monodispersity and Absence of Self-Association
Phenomena
This was supported by (i) the observation of only single boundaries from sedimentation
velocity (Figures 1a and 2a), (ii) no evidence of an increase in sedimentation
coefficients with increase in concentration (Figures 1b and 2b), and (iii) linear plots of
log (absorbance) versus distance squared (Figures 3b and 4b) from sedimentation
equilibrium.
Figure 1a. Sedimentation velocity profiles for Fab'.
Loading concentration = 2.0 mg/mL; rotor speed = 49,000 rpm; scan interval = 27 min. The
direction of sedimentation is from left to right.
Figure 1b. Plot of sedimentation coefficient versus
concentration for Fab'. Run conditions were as above.
Figure 2a. Sedimentation velocity profiles for F(ab')2.
Loading concentration = 6.6 mg/mL; rotor speed = 49,000 rpm; scan interval = 18 min. The
direction of sedimentation is from left to right.
Figure 2b. Plot of sedimentation coefficient versus
concentration for F(ab')2. Run conditions were as above.
Figure 3a. Sedimentation equilibrium of Fab'. Solute
distribution recorded at 9000 rpm. Loading concentration = 0.5 mg/mL.
Figure 3b. Plot of log (absorbance) versus Xi for Fab'.
Figure 3c. Plot of M* function versus Xi (the
normalized radial displacement parameter) for Fab'. Xi = (r2 - a2)/(b2
- a2) where r is the radial displacement and a and b are the corresponding
values at the cell meniscus and base, respectively.
Figure 4a. Sedimentation equilibrium of F(ab')2.
Solute distribution recorded at 9000 rpm. Loading concentration = 0.5 mg/mL.
Figure 4b. Plot of log (absorbance) versus Xi for
F(ab')2.
Figure 4c. Plot of M* function versus Xi (the
normalized radial displacement parameter) for F(ab')2.
Absolute Molecular Weights
The absolute molecular weights (i.e., not requiring
assumptions concerning calibration standards), M0w,
calculated from the limiting form of the function M* to the cell base (see Ref. 2) were 47,000 ± 2000 for Fab' (Figure 3c) and 94,000 ± 2000
for F(ab')2 (Fig. 4c). These are in almost exact agreement with the polypeptide
sequence molecular weights of 47,499 and 94,996, respectively. This demonstrates both the
accuracy of the Optima XL-A and the homogeneity of the antibody preparations.
Sedimentation Coefficients
The s020,w values for the Fab' and F(ab')2
fragments were determined as 3.6 ± 0.2 S and 5.1 ± 0.2 S, respectively (Figures 1b and
2b). The relationship between these values and molecular weight are in accord with the
sedimentation behavior of other globular proteins that have been well characterized.
Solution Conformation Modeling
The s020,w values so found can be used as a
useful handle on the conformation of these molecules in solution. Using simple formulae,(6) a combination of the s020,w
and molecular weight data yields estimates for the frictional ratios (f/f0)
of 1.31 ± 0.11 and 1.47 ± 0.08, respectively. These values can be interpreted in terms
of simple ellipsoid models (see Ref. 7), or, more usefully, it
is possible to use this data to model conformation in terms of sophisticated hydrodynamic
bead models using the FORTRAN program TRV.(8)
However, this increased sophistication requires additional data from other solution
measurements, such as x-ray scattering. For example, the hydrodynamic bead model shown in
Figure 5 for Fab' was constructed on the basis of an initial estimate for the solution
conformation based on static x-ray crystallographic coordinates(9)
for the Fab' fragment of monoclonal antibody R19.9 and subsequently modified in order to
reproduce f/f0 (or equivalently s020,w) and
a radius of gyration of 26 ± 3 Å obtained from small angle x-ray scattering experiments.
This data is currently being extended by us to model the conformation of F(ab')2,
particularly in terms of antigen binding site separation, with the ultimate aim of
adequately representing the solution conformation of the intact parent antibody.
Figure 5. Hydrodynamic bead model for Fab' of chimeric
B72.3 (Celltech Ltd), based on sedimentation coefficient and radius of gyration data. The
sphere coordinates and radii were generally such that their maximum diameter is 10 Å. The
overall dimensions of this model are approximately 50 x 80 x 50 Å, inclusive of
hydration.
References
- Begent, R. H. J. Recent advances in tumour imaging. Use of radiolabeled
antitumour antibodies. Biochim. Biophys. Acta 780, 151-166 (1985)
- Creeth, J. M., Harding, S. E. Some observations on a new type of point
average molecular weight. J. Biochem. Biophys. Methods 7, 25-34 (1982)
- Yphantis, D. A. Equilibrium ultracentrifugation of dilute solutions. Biochemistry
3, 297-317 (1964)
- Harding, S. E., Horton, J. C., Morgan, P. J. MSTAR: a FORTRAN program
for the model independent molecular weight analysis of macromolecules using low or high
speed sedimentation equilibrium. Analytical Ultracentrifugation in Biochemistry and
Polymer Science, Chapter 15. Edited by S. E. Harding, J. C. Horton and A. J. Rowe.
Cambridge, Royal Society of Chemistry, (in press).
- Perkins, S. J. Protein volumes and hydration effects. The calculations
of partial specific volumes, neutron scattering matchpoints and 280-nm absorption
coefficients for proteins and glycoproteins from amino acid sequences. Eur. J. Biochem.
157, 169-180 (1986)
- van Holde, K. E. Sedimentation. Physical Biochemistry, pp.
110-121. 2nd ed. Englewood Cliffs, N. J., Prentice Hall, 1985.
- Harding. S. E. Modelling the gross conformation of assemblies using
hydrodynamics: the whole body approach. Dynamic Properties of Biomolecular Assemblies,
pp. 32-56. Edited by S. E. Harding and A. J. Rowe. Cambridge, Royal Society of Chemistry,
1989.
- de la Torre, J. G. Hydrodynamic properties of macromolecular assemblies.
Dynamic Properties of Biomolecular Assemblies, pp. 3-31. Edited by S. E. Harding
and A. J. Rowe. Cambridge, Royal Society of Chemistry, 1989.
- Lascombe, M. B., Alzari, P. M., Boulot, G., Saludjian, P., Tougard, P.,
Berek, C., Haba, S., Rosen, E. M., Nisonoff, A., Poljak, R. J. Three-dimensional
structure of Fab R19.9, a monoclonal murine antibody specific for the p-azobenzenearsonate
group. Proc. Natl. Acad. Sci. 86, 607-611 (1989)
|
