Crenezumab, as shown in the figure above,
lowers Abeta oligomers levels in CSF. The figure shows boxplots of Abeta oligomer levels at baseline and week 69 (WK69) of crenezumab
treatment. Dots represent mean levels of the Abeta oligomer concentration from matched CSF samples of individual AD subjects. Samples
with values below the lower limit of quantitation (LLoQ) are shown in red. Boxes indicate 25th to 75th percentile;
horizontal bar indicates median.
As shown above, the KD of crenezumab for
Abeta oligomers is in the picomolar range (0.4 - 0.6 nM) while for the monomeric form of Abeta, the antibody has a comparatively
faster off rate resulting in an overall ~10-fold lower affinity (3.0 - 5.0 nM; Atwal et al., ADPD 2017 presentation, Ultsch et
al., 2016). The binding preference for oligomeric forms of Abeta measured in vitro, translates into a significant reduction of
Abeta oligomers in CSF of AD patients treated with crenezumab (ABBY and BLAZE Phase 2 trials), where 86% of patients dosed intravenously
(IV) and 89% of patients dosed subcutaneously (SC) display lower levels of CSF Abeta oligomers at week 69 than at baseline (p<0.01
for IV and p<0.001 for SC vs. placebo; Yang/Selkoe, AAIC 2018 presentation; Figure 10). These data provide strong evidence that
the principal targets, engaged by crenezumab in the CNS of AD subjects, are Abeta oligomers.
Significance of crenezumab’s
To describe the Abeta-crenezumab interaction
with atomic resolution, the crystal structure of crenezumab (as a Fab fragment) in complex with Abeta 11-25 was resolved at 2.3 Å
(Ultsch et. al., 2016). The structure reveals a well-defined contiguous interaction between crenezumab and Abeta residues His13-Val24,
in an extended conformation.
Figure 11: The crystal structure of
Ref: Ultsch, et. al., Sci Rep 2016
The crystal structure of crenezumab (Fab’)
shown above complexed with Abeta11-25 peptide. Crenezumab binds and sequesters the hydrophobic core of Abeta breaking a salt-bridge
characteristic and essential for the formation of the beta-hairpin conformation, eliminating key features of the basic organization
in Abeta oligomers and fibrils. Green mesh shows the electron density map corresponding to the Abeta peptide.
The observed binding mode is consistent
with high affinity for multiple forms of Abeta, explaining crenezumab’s binding to a range of Abeta species, particularly
to Abeta oligomers on a molecular level. The conformational requirements for epitope recognition includes the following subtle
but critical element that is likely the basis for crenezumab’s versatile binding profile and suggestive of the therapeutic
mechanism of action: binding of crenezumab to Abeta breaks a salt-bridge hairpin turn essential for Abeta oligomer formation between
Asp23 (i.e., within the mapped epitope) and Lys28 located in the main hydrophobic segment of Abeta (Figure 12).
Binding of crenezumab to the central epitope
within the core of the toxic amyloid beta-sheet assembly explains the observed inhibition of Abeta aggregate formation, as well
as the disaggregation propensity of pre-formed Abeta fibrils. Using in vitro aggregation assays, this anti-aggregation activity
of crenezumab was greater than the one of an antibody binding to an N-terminal epitope.
The below Abeta salt-bridge hairpin turn
is responsible for the self-association and subsequent oligomerization into toxic beta-sheet conformations. Due to the orientation
of the heavy-chain residues, crenezumab binds to Abeta structures compatible with the hairpin-like turn, but not an alpha-helix.
As only Abeta monomers, but not oligomers or aggregates, can adopt an alpha-helical structure, this likely explains why crenezumab
favors interaction with oligomeric over monomeric Abeta.