BONE-TARGETED ALKALINE PHOSPHATASE - Fusion Protein Technologies for Biopharmaceuticals: Applications and Challenges - page 415

FIGURE 28.1 (A) Three-dimensional modeling of ENB-0040, constructed by Dr. Boguslaw Stec,
Sanford-BurnhamMedical Research Institute, La Jolla, CA. The sequence of the whole construct was
analyzed. The sequences of fragments corresponding to particular molecular elements constituting
the ENB-0040 molecule were submitted to the FFAS03, the molecular similarity searches against the
structures present in the PDB (1ZED for alkaline phosphatase and 1HZH and 1IGZ for the antibody
moiety). As a result, the closest sequence analogs of individual fragments, that is, the alkaline
phosphatase, the Fc antibody fragment were selected. The full sequence models were created using
the model building tools present in FFAS03 (Modeller) [68]. Program was used to create these
molecular fragments. Subsequently, the models were positioned against each other in Coot
(the crystallographic modeling software) in order to make required necessary connections to create
the whole molecule. The linkers as well as the C-terminal Asp-rich regions were modeled by inserting
the missing sequences in the extended conformation for the linker and helical conformation for the
poly-Asp tail. Finally, the structure with optimized double disulfide bridge was energy minimized in
Sybyl. The model shows rigid AP and Fc modules connected by a highly flexible linker reminiscent of
the neck area of the mature antibody. Two disulfides bonds are marked by spheres. The terminal poly-
Asp region is exposed on the opposite site of the AP module. The whole structure is dimeric that
conforms to the preferred oligomeric state of the AP as well as the Fc region of the antibody. The
figures and final visualization was performed in Pymol. (B) The affinity of the purified ENB-0040 for
hydroxyapatite mineral was compared to that of soluble TNAP purified from bovine kidney. Total
activity was the sum of enzymatic activity recovered in both free and bound fractions and was found
to be 84% and 96% of the spiked enzymatic activity for the bovine and ENB-0040 forms of enzyme,
respectively. ENB-0040 bound 32-fold more efficiently to reconstituted hydroxyapatite (HA) than
did unmodified bovine kidney TNAP. Furthermore, most recombinant ENB-0040 utilized in the assay
was accounted for by summing enzymatic activity recovered in bound and nonbound fractions,
indicating that ENB-0040 binding to mineral does not significantly alter its enzymatic activity .
(C) Histochemical staining for ALP activity in long bones of an ENB-0040-treated Akp2 / mouse
compared with an age-matched untreated Akp2 / mouse. (Data taken from reference [23]).

protein for hydroxyapatite mineral was contrasted to that of
soluble TNAP derived from bovine kidney. ENB-0040 binds
32-fold more efficiently to reconstituted hydroxyapatite than
does bovine kidney TNAP (Figure 28.1B). Furthermore,
most of the recombinant ENB-0040 protein introduced in
the assay could be accounted for by summing enzymatic
activity recovered in both the bound and nonbound fractions
[23]. This analysis indicated that ENB-0040 binding to
mineral does not significantly alter its enzymatic activity.

The ENB-0040 product is a sterile, colorless, aqueous
solution with no visible particulates and is available at either
40mg/mL or 100mg/mL drug in 25mM sodium phosphate
buffer, pH 7.4, and 150mM sodium chloride, packaged in a
2mL glass vial. The product is dispensed in an extractable
volume of 0.5mL, without additional formulation. Investi-
gational supplies for clinical use should be stored refriger-
ated from 2 to 8 C. ENB-0040 is compatible with infusion
bags at 19-23 C for up to 2 h.