Information Technology Reference
In-Depth Information
MacroModel (Schrodinger, Inc.), and MOE (Chemical Computing Group, Inc.) are
commercially available. Additional modeling programs are freely available on the
Internet; though not as complete as the commercially available products, they per-
form many of the same functions. Such programs include pymol
(http://www.pymol.org) for visualization and in silico mutagenesis of proteins,
namd (http://www.ks.uiuc.edu) for molecular dynamics simulations, vmd (http://
www. ks. uiuc.edu) for more advanced visualization, and grace (http://plasma-
gate.weizmann.ac.il/Grace/) for data analysis.
Using any of these programs, differences in structure between wild-type and
amino-acid-substituted protein sequences can be measured by calculating the RMS
deviation of the positions of the alpha-carbons in the protein chains after identifying
the lowest energy state of each sequence. A relatively low RMS deviation suggests
that the mutations do not cause broad structural changes in the protein, while a rela-
tively large RMS difference suggests that the substitutions do affect tertiary struc-
ture. Close examination of the results permits the exact identification of the amino
acids responsible for the greatest energy difference; thus, resubstitution and reitera-
tive analysis of the effect of substitutions on protein structure may be performed.
6.5. When Can Deimmunization Be Useful?
Deimmunization may be extremely useful at certain stages of protein therapeutic
development. Three scenarios are illustrated in the next few paragraphs.
6.5.1 Prioritizing in the Preclinical Stage of Development
Given several similar candidates that may have all demonstrated a reasonable level
of efficacy in preclinical evaluations, drug developers need a means for selecting the
one or two that are most likely to succeed. One means of reducing the list of candi-
dates to evaluate is to score each therapeutic protein on a “potential immunogenicity
scale” such as the one described in Section 2.3.3. This scale allows for comparisons
between proteins that are known to be nonimmunogenic, such as the constant regions
of human antibodies, and viral or bacterial proteins known to be highly immunogenic
such as tetanus toxin, ESAT6 derived from TB, or haemagglutinin derived from
influenza. Candidate therapeutic proteins displaying low or limited potential for
immunogenicity may be prioritized for clinical trials using this scale. Candidate
therapeutics displaying high potential immunogenicity might be set aside or returned
to the developmental pipeline for reengineering.
6.5.2 Modifying a Lead Candidate Following Immunogenicity Testing
In some cases, a good candidate has been identified but is predicted to contain an
unacceptable amount of immunogenic potential. In this case, regions of the thera-
peutic protein that account for at least 50% of the total potential for immunogenic-
ity contained within the candidate protein can be identified. These regions of high
potential can then be evaluated in order to identify individual amino acids that are
