Page 177 - Read Online

Page 4 of 23 Padarti et al. Vessel Plus 2018;2:21 I http://dx.doi.org/10.20517/2574-1209.2018.34

CCM1 NUDIX domain
The NUDIX domain is found in the N-terminus (residues 1-170) of CCM1 and it contains a stretch of basic
residues that potentially interacts with microtubules. NUDIX domains are usually found in hydrolases that
bind to a variety of substrates. The NUDIX fold in CCM1 is a centrally positioned β-sheet with flanking
α-helices. Traditionally, the NUDIX domain contains a NUDIX box which contains the Gx Ex REUxEExGU
5
7
motif . However, CCM1 doesn’t contain a traditional NUDIX motif, yet the tertiary structure of N-terminus
[53]
of CCM1 still adopts a NUDIX fold. The catalytic residues found in other NUDIX domains are missing in
the CCM1 NUDIX domain . Therefore, the function of the NUDIX domain in CCM1 may be different
[54]
from other NUDIX domain-containing proteins. Despite the similarities, superimposition of the X-ray
crystal crystallography of the NUDIX domain with known substrates (81 in total) revealed no potential
binding partners. Therefore, the function of this domain is still uncertain . However, sequence analysis
[54]
has elucidated the presence of several known sequences such as potential tubulin binding sequence and a
[48]
nuclear localization sequence (NLS) within the NUDIX domain.
[42]
CCM2 is the second largest of the CCM proteins, 444 amino acids in length and contains a PTB domain
at the N-terminus and a harmonin homology (HH) domain at the C-terminus . 20% of all familial forms
[3]
of CCM are due to mutations of CCM2 [36,39] , however, the penetrance of CCM2 mutation was reported to
be 100% . CCM2 is found ubiquitously expressed in the endothelial cells (EC) from various organs [55,56] .
[40]
Despite the lack of a recognizable NLS and NES, CCM2 is found in both the nucleus and cytoplasm due to its
interaction with CCM1 [18,23,57] . In the absence of functional CCM1, CCM2 is not localized to the cell junction,
however, this function is recovered with the addition of wild-type CCM1, implying that binding to CCM1 is
essential for localization of CCM2 to the cell junction .
[58]
CCM2 HH domain
CCM2 functions as the scaffold in the CSC with binding sites for both CCM1 and CCM3 . The HH domain
[16]
at the C-terminus (residues 283-375) is comprised of 6 packed α-helices termed H1*, H1, H2, H3, H4, and
H5 in that order from N-terminus to C-terminus. The H1* is a short α-helix with 3 amino acid residues and
H4 is a 3 helix that contains 13 residues. This domain is stabilized by several intramolecular interactions
10
(i.e., R346 to E314, R354 to E366, and P355 to F356). This C-terminal domain bears structural similarity to
harmonin protein and therefore termed HH domain. Although there is structural similarity, CCM2 HH
domain is able to bind to neither Cadherin 23 (a validated binding partner of Harmonin) nor to CCM3. This
HH domain exists in two conformations, monomeric and dimeric. The dimeric form has an increased affinity
for dimerization; however, there are no sufficient data to affirm the occurrence of dimeric CCM2 in vivo yet .
[45]

CCM2 PTB domain
The N-terminus of the CCM2 contains a DAB-like PTB domain . This domain contains 2 β-sheets
[59]
composed of 7 β-strands, with α-helices capped at both ends. It was shown that this PTB domain binds to
the NPXY motifs present in CCM1. Yeast two-hybrid assays showed that CCM2 PTB domain was able to
bind to a CCM1 construct that contained the second and third NPXY motifs but not the first motif [17,18] .

CCM3 is the smallest of the 3 CCM proteins with 212 amino acids. CCM3 mutations tend to result in the
most aggressive form of the disease . 10% of all familial forms of CCM are due to mutations in CCM3
[8]
gene [36,39] . The penetrance of CCM3 mutation is approximately over 60% . In one study that sought for
[40]
promoter variants for the CCM genes, two protective single nucleotide polymorphisms were identified in
the promoter region of CCM3 (rs9853967 and rs11714980) to be associated with CCMs, while no causative
variants were identified in the promoter regions of CCM1 or CCM2, among the selected CCM patient cohort.
These variants could partially explain the range of disease burden seen in CCM . CCM3 is localized to the
[60]
cis-face of the Golgi body . Its interactions with phospholipids, PtdIns , were believed to facilitate the
[61]
[3-5]
translocation of CCM3 to the plasma membrane . CCM3 is a 2-domain-containing protein where both
[62]
domains are conjoined by a flexible hinge region . The tertiary structure of CCM3 is a V-shaped structure.
[63]