Chemistry Reference
In-Depth Information
principle, a domain has a compact three-dimensional structure that may fold
independently of the rest of the protein, and may function and evolve semi- inde-
pendently. The sequence of a protein domain is typically 100-200 amino acids
long. A given CTLDcp by defi nition contains an operative lectin domain, but in
addition may contain multiple diverged copies of it with different sugar or other
ligand - binding specifi city as well as other domains. These features are shown in
Figure 20.2 .
The development of the C-type lectin fi eld benefi ted greatly from the early efforts
of researchers, notably Kurt Drickamer, to systematize the disparate biological data
[1, 2]. This led to their characterization as lectins with CRDs of length 110-140
residues, which bound carbohydrates in a Ca
2+
- dependent manner. Furthermore,
alignment of their sequences showed this protein domain had a characteristic
conserved sequence signature. This feature has been of enormous value in advanc-
ing C-type lectin research as it has allowed initial identifi cation as a potential
CTLDcp directly from analysis of the protein sequence. A further helpful fi nding
from this early sequence analysis was that carbohydrate specifi city is often corre-
lated with a particular tripeptide sequence motif within the sequence signature -
E
P
N
(Glu - Pro - Asn)
for
mannose -type ligands and
Q
P
D
(Gln - Pro - Asp)
for
galactose-type ligands. This permits initial functional predictions.
Understanding of the structural basis of the sequence signature came with solu-
tion of the fi rst X - ray structure of a CTLD - of rat mannan - binding protein ( MBP ) -
A - in 1991 and a little later by a structure of this CTLD with a bound mannose
molecule. Comparison of the residues then identifi ed as the sequence signa-
ture - 12 totally conserved and 18 conservatively conserved - against the three -
dimensional structure allowed the roles for most of them in Ca
2+
-
and
carbohydrate-binding and stabilizing the protein fold to be defi ned [1] .
The basic protein fold is shown in Figure 20.1 and the main details of carbohy-
drate binding in Figure 20.3. The key characteristics of the CTLD fold are two
antiparallel
- helices, and two disulfi de bridges and a hydropho-
bic core stabilizing the long-loop region which contains the primary sugar-binding
site. Examination of the large number of CTLD structures that have been obtained
by crystallography shows that this unique loop- in - a - loop structure, in which the
large fl exible long-loop region is maintained on a stable core, allows the fold to
tolerate substantial variation in the shape of the primary ligand-binding site and
adjacent regions [4]. This allows specifi c binding of large multivalent ligands such
as complex-type oligosaccharides and mannose-rich structures (triantennary
N
-glycans and mannans are examples given in Chapter 19 ), non - carbohydrate
ligands, and even both. Formation of quaternary complexes, for example in the
trimers of the group III vertebrate collectins (for explanation of this term, see
Chapter 19), further increases both specifi city and affi nity of carbohydrate recogni-
tion by C- type lectins [3] .
In 1993, Drickamer classifi ed mammalian CTLDcps into seven groups (I- VII).
This was based primarily on their domain architecture, but the grouping also
appeared to refl ect evolutionary history as it correlated well with the results of
phylogenetic analysis of the CTLD sequences [1] . The classifi cation was revised in
β
- sheets and two
α
