The discovery of AQP1 in 1992 by Agre and coworkers(Prestos et at, 1991, 1992) and
subsequent designation of "aquaporin" have opened a new field in membrane transport
physiology (Agre P et al,1993). The original definition of aquaporin was "the MIP26
homologue proteins which rapidly and selectively permeate water". MIP26 is the major
intrinsic protein of mammalian lens cloned in 1984 (Gorin et al, 1984), and many
homologous proteins since discovered make a large and growing family of membrane
integral proteins, the MIP family (Park et al, 1996). Thus, a more simple definition of
aquaporin is "MIP family members whose water channel function is confirmed". The
basic structure of MIP proteins is summarized in Fig. 1. Each molecule consists of 6
transmembrane domains, with the aminoand carboxy-terminal ends located in the cell
interior. The first half and the last half of the sequences are homologous (tandem
repeats), indicating the occurrence of gene duplication during the evolution. Each half
has the conserved motif of asparagine-proline-alanine (NPA box). Most MIP proteins
are small proteins which consist of 260¡300 amino acids. As MIP family members can
be easily cloned by PCR- based cloning strategy using the conserved NPA motif, their
numbers have drastically increased. Most Cloned MIP proteins have not been examined
for their water channel function, and thus they remain categorized as MIP proteins.
Some of them which have been proven to be water channel, thereby have acquired a
new name, "aquaporin" Some of MIP proteins have been shown not to be permeable to
water, but instead permeable to small nonionic solutes such as glycerol and urea. These
proteins are not called aquaporin. Accordingly, there are fewer aquaporins than MIP
proteins. Nonetheless, most MIP proteins may still be proven to be aquaporins if they
are adequately examined for their water channel function. For example, MIP26 was
originally thought to be a non-selective ion channel, but recent careful expression study
in Xenopus oocytes has shown its small water channel function. Thus MIP26 is now
called AQP0. Many research in this field and genome projects in many organisms have
identified plenty of MIP proteins (Park et al,1996). E coli has two MIP proteins with a
distinct functional difference; one is an aquaporin (AqpZ) (Calamita et al,1995) and the
other is a glycerol facilitator (GlpF). Like bacteria, other organisms have many MIP
proteins. For example, C. elegans has 8 and yeast has at least 4 members (Park et
al,1996, Ishibashi et al, in press). There are many MIP proteins in the plant kingdom
(for example, Arabidopsis has more than 23 members). So far 10 aquaporins have been
reported for mammals (Fig.2). Thus, the number of identified MIP members has
increased drastically, and the prevalence of MIP/AQP proteins in almost all organisms
clearly indicates their indispensable roles in the body, possibly as water and small
neutral solute transporters. The phylogenetic analysis of mammalian aquaporins shows
that they can be separated into two subgroups. This sequence homology-based
separation can be applicable to the entire MIP/AQP family and detailed discussion for
this is available in elsewhere (Ishibash et al, in press, Sasaki et al,1998). The upper
group consists of 7 aquaporins; AQP0, 1, 2, 4, 5, 6, 8. and the lower group consists of
3; AQP3, 7, 9. Functional studies have indicated that the upper members permeate water
selectively, while the lower members permeate small solutes such as glycerol and urea.
This functional difference may come from their prototypes in E. coli. One MIP protein in
E. coli. is AqpZ and function as an aquaporin with no permeability to other solutes
(Calamita et al, 1995), The other is a glycerol facilitator (GlpF) which permeates
glycerol and excludes water. Our recent study comparing functional characteristics of
AQP2 and AQP3 has indicated that pore structure of the two groups is the same and
water and small solutes pass through the same pore (Kuwahara et al, 1997).
Identification of the mechanism of selectivity of the pore of MIP/AQP proteins would be
a big challenge in this field of research. Comparison of these two subgroups may
provide some important insights. Recent advance in an electron crystallography will be
of great help for this purpose (Walz et al, 1997).
|