ArfGAPs: Not Only for the Termination - Ras Superfamily Small G Proteins: Biology and Mechanisms 2 - page 260

curved membranes. The binding site for COP-I on the Arf1 molecule does not
overlap with that of ArfGAP1 (Goldberg 1999 ). Via the ALPS motif, ArfGAP1 is
then recruited by GTP-Arf1 to the COP-I-coated curved membranes, and hydro-
lyzes GTP-Arf1. This hydrolysis was previously thought to release the COP-I coat
from vesicles (Tanigawa et al. 1993 ). However, it was shown later on that COP-I
remains on the membrane even after GTP-Arf1 is hydrolyzed and dissociated from
the membrane (Presley et al. 2002 ). Therefore, dissociation of COP-I from vesicles
may not necessarily be linked to the hydrolysis of GTP-Arf1 and hence the release
of Arf1 from the vesicles. Consistently, a proteomic analysis of purified COP-I-
coated vesicles showed that Arf-GTPases are rarely found within the COP-I
vesicles (Gilchrist et al. 2006 ) (Fig. 11.2 ). Therefore, it is likely that some ArfGAPs
may inactivate their cognate Arfs at each location of membrane curvature, even
before the overall vesicular structures are formed and vesicles are separated from
the donor membrane.
On the other hand, AMAP/ASAP- and ACAP-family members contain the BAR
domain, which may sense the membrane curvature, as well as may generate,
stabilize, and stably bind to curved membranes, as also mentioned earlier. It was
proposed that ASAP1 plays a role in the formation of curved membranes:
GTP-Arf1 is anchored to the membrane via its myristoylation, and the binding of
ASAP1 to GTP-Arf1 then vents the Arf1-containing membrane (Nie et al. 2006 ).
Consistent with this notion, the BAR domain of ASAP3/ACAP4 was recently
demonstrated to have the potential to tubulate liposomes in vitro (Zhao
et al. 2013 ). Therefore, it is plausible that ASAP1 has a dual function: one is to
promote vesicle formation mediated by GTP-Arf1, and the other is to hydrolyze
GTP-Arf1. The GAP activity of ASAP1 toward Arf1 requires phosphatidylinositol-
phosphates, such as PI(4,5)P 2 (Kam et al. 2000 ). Therefore, it is plausible to assume
that ASAP1 hydrolyzes GTP-Arf1 to release Arf1 from the vesicles, only after
sufficient amounts of modified lipids are accumulated.
A recent study on the co-crystal structure of the ASAP3 and Arf6 complex
revealed that Ca 2+ promotes the GAP activity of ASAP3 toward GTP-Arf6 (Ismail
et al. 2010 ). ASAP3 is an AMAP/ASAP-family member; and Gln479 of ASAP3,
crucial for Ca 2+ binding, is conserved among the AMAP/ASAP-family members
and also in ARAP3. This notion was intriguing because we have shown that
AMAP1 and AMAP2, members of the ASAP family, bind stably to GTP-Arf6,
even in the presence of Mg 2+ without hydrolyzing GTP, and have proposed that
these ArfGAPs act to recruit proteins, other than coatomers, to the sites of Arf6
activation, as mentioned earlier. It can hence be hypothesized that these ArfGAPs,
as far as Ca 2+ is not supplied, act as (1) binding proteins to GTP-Arf6 in order to
recruit certain proteins to the sites of Arf6 activation, and simultaneously act as
(2) coatomer components for Arf6-mediated vesicles to vent the membrane; and
that Ca 2+ spikes, perhaps induced by various extracellular stimuli, then trigger
hydrolysis of GTP-Arf6 and hence strip the vesicles from these ArfGAPs
(Fig. 11.3 ). In this regard, it should be noted that GDP-Arf6 is known to remain
bound to membranes primarily via its myristoylation and its N-terminal basic helix,
while GDP-Arf1 is no more bound to membranes even when it is still myristoylated