Biomedical Engineering Reference
In-Depth Information
computer program ROC was utilized in this design. The means to achieving
the design were: (i) shortening helix sequence length (to 20 residues, which
is the typical helical length in 3-helix bundles and thus shorter than in coiled
coils); (ii) incorporating loops with N-cap motifs; and (iii) incorporating
charges to electrostatically favour desired helix-helix pairings and disfa-
vour others. Interfacial Glu, Arg and Lys side chains were important.
A counterclockwise topology was observed by NMR. However, NMR
relaxation studies revealed a higher degree of fluidity in the core of a 3 D
than that observed in natural proteins of similar size [75].
Furthermore, a helix-binding 3-helix structure was developed by
DeGrado, Lombardi and coworkers [76]. They also developed a 3-helix
structure in which the helices were linked by disulfide bridges [77].
Pecoraro and coworkers have developed a series of triple-helical struc-
tures, self-assembled from Cys-containing homotrimers, which together
bind Hg(II) [78,79]. The starting point provided the TRI sequence, which
does not contain Cys and which forms a trimer. They also mutated
CoilV a L d (see above) into CoilV a L d -V16C. While TRI can be both
parallel and antiparallel, CoilV a L d was designed to be parallel.
6.3.3.2
4-helix bundles
The purposely simplistic a 1 A and a 1 B sequences (Table 6.1) by Ho and
DeGrado were reported to form tetrameric assemblies as 4-helix bundles
according to size-exclusion chromatography (and concentration-depen-
dent CD) [80]. Starting from these single-helix peptides, which self-
assemble weakly to tetramers, Ho and DeGrado designed helix-loop-
helix sequences that dimerize to 4-helix bundle assemblies. These con-
structs, a 1 B-Pro-a 1 B and a 1 B-Pro-Arg-Arg-a 1 B, form relatively stable
dimers in solution. The latter contained Pro-Arg-Arg as a loop, which
was designed to favour antiparallel dimerization by repulsion of electro-
static charges. This loop was eventually developed further to the useful
sequence GPRRG.
In helix-loop-helix sequences that were designed to dimerize to form
4-helix bundles, DeGrado and coworkers had previously observed a
dynamic character of the de novo -designed protein a 2 B. a 2 B had only
Leu, Glu and Lys in the sequence and an all-Leu core; its dynamic char-
acter was most likely due to rapid motions of side chains in the bundle and
the formation of multiple, interconverting topologies [2,3]. They designed
several 35 AA helix-loop-helix sequences (N-terminus acetylated,
C-terminus amidated) intended to dimerize. The GPRRG loop causes
Search WWH ::




Custom Search