Encyclopedia of Signaling Molecules

2018 Edition
| Editors: Sangdun Choi

Aquaporin

Reference work entry
DOI: https://doi.org/10.1007/978-3-319-67199-4_101692

Synonyms

Historical Background

Water transport across cell membranes is an essential requirement for the biochemistry of life, yet it took more than two centuries of observations and research before transmembrane water channels termed aquaporins were discovered. For much of this time, it was assumed that water transport occurred either via simple diffusion or in certain instances was powered by active processes. As early as the eighteenth century, it was noted that adult frogs and toads absorb water across their ventral skin when dehydrated and placed in water. A century and a quarter later, it was discovered that the increased absorption of water across the skin could be enhanced by the injection of neurohypophysial extract (pituitrin), which was known to have antidiuretic effects in mammals. These observations were subsequently termed the water-balancing effect or Brunn reaction (Heller 1945) and paved the way for mechanistic studies of water transport in frog skin. Indeed, it was by comparing the diffusion in isolated frog skin to the in vivo osmotic water uptake of the intact animal that the first experimental evidence for facilitated water transport emerged in the laboratory of August Krogh (Hevesy et al. 1935). Using heavy water (D2O), it was noted that the abdominal and thigh skin of the edible frog (Rana esculenta) was bidirectionally much more permeable to D2O than skin isolated from other areas, and that the in vivo osmotic uptake of ordinary water (H20) was three- to fivefold higher than that expected from diffusion and the osmotic gradient between the blood and external water (Hevesy et al. 1935).

Subsequent experiments revealed that the Brunn reaction could be elicited in the isolated ventral skin of anurans, but in contrast to the notion that water must be actively transported in the distal renal tubules of Mammalia to promote antidiuresis, the experiments on frog skin implied that the increased water flux associated with the Brunn reaction is a passive process involving an increase in the permeability of the skin. By incorporating the Brunn reaction in a repeat of Hevesy, Hofer and Krogh’s (1935) experiments and analyzing the results with what became known as the flux-ratio theorem, Koefoed-Johnson and Ussing (1953) confirmed the earlier results and proposed that “the response to neurohypophysial extract of the water flow through the anuran skin may consist in an opening up of pores or creation of new pores to allow a more ready flow of solution” (1953). This was the first suggestion that water pores or channels should mediate the passive transport of water through biological membranes. Physiological studies of mammalian red blood cells further supported the existence of “water-filled pores in the membrane” (Paganelli and Solomon 1957), which could be reversibly closed in the presence of mercurials or excess cysteine that bind buried sulfhydryl groups in the pores (Macey and Farmer 1970). These latter studies lead the authors to observe “that water channels transport water and very little else” (Macey and Farmer 1970). Although the first visual evidence of such elusive pores was attained through electron microscopic observations of protein aggregates appearing in the urinary bladder membranes of toads in response to vasopressin, it was the isolation of proteins from the membranes of human red blood cells that eventually led to the discovery and proof of aquaporins.

With hindsight it is now known that the very first aquaporin to be sequenced and structurally analyzed was isolated from rat and calf lenses and termed the major intrinsic polypeptide (MIP) of lens membrane; however, although the predicted structure of this protein was consistent with an aqueous channel with six membrane-spanning domains and intracellular amino- and carboxy-termini, its primary function was considered as a gap junction protein (Gorin et al. 1984). Parallel studies conducted on human erythrocytes, however, took advantage of earlier observations that mercurials were associated with closure of the putative pores. These latter studies succeeded in isolating proteins (band 3, 80–100 kDa and band 4.5, 35–60 kDa) that bound a mercury compound ([203Hg]-p-(ch1oromercuri)benzenesulfonate, PCMBS) and were correlated to the inhibition of water influx as measured by nuclear magnetic resonance (Benga et al. 1986). Although both bands had been associated with anion or glucose transport, respectively, it was suggested that “It remains possible that a minor membrane protein that binds PCMBS is involved in water transport” (Benga et al. 1986). By chance, a 28 kDa integral membrane protein was indeed isolated from human erythrocytes and renal tubules by other investigators (Denker et al. 1988) but was initially considered to be a breakdown product of Rh(D) blood group antigens. The 28 kDa peptide nevertheless turned out to represent a nonglysoylated form of the 35–60 kDa higher molecular weight (HMW 28 kDa)-protein (Denker et al. 1988) and was termed channel-like integral membrane protein of 28 kDa (CHIP28) following initial peptide sequencing and isolation of the cDNA. The structural features of the deduced protein closely resembled that of bovine MIP, and subsequent heterologous expression in frog oocytes followed by hyposmotic shock demonstrated for the first time the water transporting property of the channel (Preston et al. 1992). Evolutionary studies had already established that MIP-related proteins are found in bacteria (GlpF), plants (TIP, NOD26), insects (Bib), and mammals (MIP) (Pao et al. 1991), and the name “aquaporins” was proposed to encompass the protein family (Agre et al. 1993). CHIP28, which represented the first functionally defined water channel, was renamed AQP1, and MIP, which is also a water channel but which also harbors cell-to-cell adhesive properties, was renamed AQP0. Ten years later, and more than two centuries after the observations of water uptake in frogs, the Nobel Prize for Chemistry was awarded to Peter Agre for his work in the discovery of aquaporins. Since then, the field has expanded exponentially with >100,000 related nucleotides now sequenced from all domains of life.

Structure

Available crystallographic data (Table 1) reveal that the secondary and tertiary structures of aquaporins are well conserved and reflect the domains first proposed for the AQP0 channel (Gorin et al. 1984). The structure of a generic aquaporin thus consists of a central pore surrounded by a right-handed bundle of six α-helical transmembrane domains linked by five loops (A–E), of which loops A, C, and E are extracellular, and loops B and E are intracellular (Fig. 1). The amino- and carboxy-terminal domains, which have the greatest structural diversity in terms of the number and identity of residues, are also located on the cytoplasmic side of the membrane. The extracellular and intracellular domains are the major sites of posttranslational modifications including glycosylation or protonation and phosphorylation associated with gating and trafficking (see section on “Regulation” below). The repetitive nature of transmembrane domains 1–3 and 4–6, which form two hemipores, indicates that the channels arose by internal duplication (Pao et al. 1991).
Aquaporin, Table 1

Aquaporin crystallographic data in Prokaryota and Eukaryota

Group

PDB

Ortholog

Resolution (Å)

Organism

Archaea

2F2B, 2EVU

AqpM

1.68–2.30

Methanothermobacter marburgensis

3NE2

AqpM

3.00

Archaeoglobus fulgidus

Bacteria

1FX8

GlpF

2.20

Escherichia coli

1LDA, 1LDF, 1LDI

GlpF

2.10–2.80

Escherichia coli

1RC2

AqpZ

2.50

Escherichia coli

2ABM

AqpZ

3.20

Escherichia coli

2O9D-G

AqpZ

1.90

Escherichia coli

3NKA, 3NKC, 3NK5

AqpZ

2.40–3.00

Escherichia coli

Plantae

2B5F, 1Z98

PIP2;1

2.10–3.90

Spinacia oleracea

3CN5, 3CN6, 3CLL

PIP2;1

2.05–2.95

Spinacia oleracea

4IA4

PIP2;1

3.10

Spinacia oleracea

4JC6

PIP2;1

2.15

Spinacia oleracea

5I32

TIP2;1

1.18

Arabidopsis thaliana

Plasmodium

3CO2

PfAqp

2.05

Plasmodium falciparum

Fungi

2W2E, 2W1P

Aqy1

1.15–1.40

Pichia pastoris

5BN2

Aqy1

1.30

Pichia pastoris

3ZOJ

Aqy1

0.88

Saccharomyces cerevisiae

Mammalia

1SOR

AQP0

3.00

Ovis aries

3M9I

AQP0

2.50

Ovis aries

3 J41

AQP0

25.00

Ovis aries I Homo sapiens

1YMG

AQP0

2.24

Bos taurus

2B6O, 2B6P

AQP0

2.40

Bos taurus

2C32

AQP0

7.01

Bos taurus

1FQY

AQP1

3.80

Homo sapiens

1IH5

AQP1

3.70

Homo sapiens

1H6I

AQP1

3.80

Homo sapiens

4CSK

AQP1

3.28

Homo sapiens

1J4N

AQP1

2.20

Bos taurus

4NEF

AQP2

2.75

Homo sapiens

2D57

AQP4

3.20

Rattus norvegicus

2ZZ9

AQP4

2.80

Rattus norvegicus

3D9S

AQP5

2.00

Homo sapiens

5DYE, 5C5X

AQP5

2.60–3.50

Homo sapiens

Aquaporin, Fig. 1

Aquaporinstructure. General topology of aquaporins showing the six transmembrane α-helices (1, 2, 4–6, 8), the two Asn-Pro-Ala (NPA) motifs, the five loops (A–E), and intracellular amino- and carboxy-terminal domains. Molecules and domains associated with posttranslational modifications, gating, and trafficking are indicated

In both prokaryotes and eukaryotes, channel protomers range in size from ~27–35 kDa, with some exceptions such as Drosophila Bib, which has an Mr of 76.5 kDa due to an extended carboxy-terminus. The eukaryote aquaporin protomers are folded in the endoplasmic reticulum and integrated into vesicular membranes mostly as homotetramers but also occasionally as heterotetramers (Fig. 2a). When folded, loops B and E retain lipophilic hemi-helices that project into the center of the molecule to present opposing and highly conserved Asn-Pro-Asp (NPA) motifs that regulate the single-file conductance of water while simultaneously functioning as a cation and a proton excluding filter (Fig. 2b). The structural features of each protomeric channel thus reflect an ancient timepiece (hourglass) that allows water or sand to pass through a central constriction. Although, highly conserved between orthologs in Prokaryota and Eukaryota, variations of the NPA motifs are found in unorthodox aquaporins encompassing gnathostome Aqp11 (e.g., NPC, NPS, NPT, NPG) and metazoan Aqp12 (e.g., NPS, NPT, CPY, NPV) and in some orthodox channels such as the first motif in trematode Aqp8-like channels (TAA), teleost Aqp3b (NPT), Aqp8ab and -8bb (NPP), and gnathostome AQP7 (NAA, NGA). In addition, several of the plant paralogs from different species display modified NPA motifs. The narrowest region of the pore is found in the outer vestibule and typically consists of aromatic residues and an arginine (ar/R) to form the major selectivity filter determining which molecular permeant can traverse the pore (Figs. 1, 2b; Table 2).
Aquaporin, Fig. 2

Aquaporin 3D structure. (a) Extracellular view of the quaternary structure of human AQP5 (3D9S) showing cartoon renderings of the four protomeric water channels (red, green, yellow, and blue) and the permeation paths of water (wheat spheres). (b) Monomeric lateral view of the water path (wheat spheres) through human AQP5 showing the aromatic-arginine (ar/R, spacefill spheres) and Asn-Pro-Ala (NPA, sticks) constrictions. (c) B-factor putty render of tetrameric sheep AQP0 (green, cyan, magenta, yellow) with interacting residues associated with cell-to-cell adhesion shown as spacefill (white). (d) Octomeric arrangement of double-layered AQP0 tetramers showing aligned interactions between residues in opposing membranes

Aquaporin, Table 2

Major solvents and solutes shown to permeate through archaeal, bacterial, protozoon, plant, fungal, nematode, arthropod, and vertebrate aquaporins. The aromatic arginine (ar/R) constriction residues are shown for each channel with positions P1 in transmembrane domain (TMD)-2, P2 in TMD5, and LE1and LE2 in loop E

Aquaporin grade

Ortholog

ar/R

Permeant

P1

P2

LE1

LE2

Classical Aqp4-related aquaporins

AQP0

F

H

A

R

Water, CO2

AQP1

F

H

C

R

Water, CO2, NO, H2O2, NH3

AQP2

F

H

C

R

Water

AQP4

F

H

A

R

Water, CO2

AQP5

F

H

C

R

Water, CO2

AQP6

F

H

C

R

Water, glycerol, urea, anions, NO3, CO2, NH3

Bib

S

C

?

R

Cations

Drip

F

H

A

R

Water

Prip

F

H

A/S

R

Water, urea

Eglp

F

A/S

A/S/G

R

Water, glycerol, urea, polyols, trehalose

PIP

F

H

T

R

Water, CO2, H2O2

Aqp8-related

AQP8

H

I

A/G

R

Water, glycerol, urea, NH3, H2O2

Nematode Aqp8L

H

I

A

R

Water

Trematode Aqp8L

A

A

C

A/S

Water

TIP

H

I

A/G

R

Water, glycerol, urea, CO2, H2O2, NH3

XIP

I/Q

A/I/T

A

R

Water, glycerol, urea, boric acid, H2O2

HIP

H

H

A

R

Unknown

Aqy1, Aqy2

F

H

A/T

R

Water

Unorthodox aquaporins

AQP11

T/P

L

A

L

Water, glycerol

Plant SIP

I/L/V

I/L/V

P

I/N

Water, metalloids, H2O2

Aquaglyceroporins

AQP3

F

G

Y

R

Water, glycerol, urea, antimonite, arsenite, polyols

AQP7

F

G

Y

R

Water, glycerol, urea, antimonite, arsenite, NH3

AQP9

F

A/S

C

R

Water, glycerol, urea, carbamides, polyols, purines, pyrimidines, antimonite, arsenite, CO2, NH3

AQP10

G

G/S

I/F/Y

R

Water, glycerol, urea; antimonite, arsenite

AQP13

F

G

Y

R

Water, glycerol, urea

Arthropod Glp

F/Y

A/T/H

A/F/T

R

Water, glycerol, urea

Nematode Glp

W

G

Y

R

Water, glycerol

Plasmodium Glp

W

V

G

R

Water, glycerol, NH3

Fps1

W

G

F

R

Water, glycerol, methylamine, NH3, antimonite, arsenite, boric acid

NIP

A/G/W

A/I/S

A/G

R

Water, glycerol, formamide, arsenite, boric acid, silicic acid, NH3

Bacteria

AqpZ

F

H

T

R

Water

GlpF

W

G

F

R

Water, glycerol, urea, antimonite, arsenite, polyols, lactate

Archaea

AqpM

F

I/V/L

A/S

R

Water, glycerol

Although the major permeants are indicated, there may be additional permeants not listed here that are transported by specific aquaporins in different organisms. “?” indicates an unclear residue. See Finn and Cerdà (2015) and Maurel et al. (2015) for references

Some aquaporins evolved the ability to form intercellular junctions, including vertebrate Aqp0, an N-terminal truncated M23 isoform of mammalian AQP4, and the Bib channel of Drosophila. While the molecular basis for the cell-to-cell adhesion of Bib remains to be determined, the structural interactions facilitating adhesion of Aqp0 and AQP4M23 channels are different. In mammalian AQP0, four proline residues (Pro38, Pro109, Pro110, Pro123) and an arginine (Arg113) in extracellular loops A and C form a triangle of residues on each protomer that interact with the same positions in the protomeric channels in the opposing membrane (Fig. 2c). The Pro38 residues on loop A form a unique rosette-like structure in the center of the octomeric junction, and the remaining residues hydrogen bond to maintain a precisely stacked double layered arrangement of the opposing tetramer (Fig. 2d). By contrast, AQP4M23 junctions are formed by Pro139 and Val142 at the extracellular end of transmembrane domain 3, which hydrogen bond to the same residues of only one of the protomers of four opposing tetramers. In this way, the junctions formed by an AQP4M23 tetramer in one membrane interact with four separate AQP4M23 tetramers in the juxtaposed membrane. Consequently, the channels are not stacked in register as for AQP0 but are arranged in a rotated lattice with each protomer partially obstructing the water path of the channel in the opposing membrane (see Wspalz et al. 2009; Fig. 4 for an illustration).

Evolution and Diversity

Evolutionary relationships of aquaporins were studied even before the water-transporting function of the proteins was established and identified evidence for the internal duplication of hemipores 1 and 2 before the divergence of prokaryotes from eukaryotes (Pao et al. 1991). Current analyses have provided evidence that prokaryotic aquaporins can be phylogenetically separated into four major grades termed AqpZ, AqpN, AqpM, and GlpF (Fig. 3a) (Finn and Cerdà 2015). The term grade is used rather than clade to accommodate the similar, but polyphyletic nature of the clusters, each of which encompasses channels present in Archaea and Bacteria. The AqpZ grade of channels are water-selective, while the AqpM and GlpF channels transport water and polyols such as glycerol. The permeability properties of the AqpN grade of channels have yet to be determined.
Aquaporin, Fig. 3

Aquaporin evolution and diversity. (a) Bayesian midpoint rooted codon tree of prokaryotic aquaporins together with plant NIPs, with posterior probabilities shown at each node. (b) Bayesian codon tree of eukaryotic aquaporins rooted with archaean AqpM, with posterior probabilities shown at each node. (c) Phylogenetic distribution of eukaryotic aquaporins arranged in four grades. N refers to the number of paralogs in the given organism

Until recently, it was thought that glycerol transporters of vascular plants, termed nodulin 26-like intrinsic proteins (NIPs), arose through horizontal gene transfer of AqpZ-type channels (Abascal et al. 2014). However, recent phylogenetic evidence suggests that bacterial AqpN-type channels were the likely origin of NIPs with the last common ancestor closely related to nitrite-oxidizing members of the Chloroflexi phylum (Fig. 3a) (Finn and Cerdà 2015).

Amongst eukaryotes, the gene copy number and diversity of channels varies considerably between the different taxonomic lineages but can be grouped into four grades of classical aquaporins related to Aqp4-type channels, Aqp8-type channels, unorthodox channels, and classical aquaglyceroporins (Fig. 3b, c). The high gene copy number in protozoan organisms is surprising, but indicates that there is no relationship between copy number and organism complexity, and may instead reflect the ability to adapt to changing environments. Evolution of the superfamily in the different taxonomic lineages is characterized by multiple gene duplications, either through whole genome duplications or local tandem duplications and subsequent differential gene loss (Abascal et al. 2014; Finn et al. 2014, 2015). One unusual case of aquaporin gene evolution has been established for insect glycerol transporters. In this instance, novel Prip-like channels, which are phylogenetically related to Aqp4 orthologs (Fig. 3b), duplicated and neofunctionalized in basal hexapods, whereby one of the ar/R residues (His on transmembrane domain 5) was replaced by smaller uncharged residues (mostly Ser and Ala). This critical alteration seems to have opened the pore to allow the passage of glycerol and other small uncharged solutes. The increased efficiency of glycerol transport and positive selection of these hexapod-specific channels, termed entomoglyceroporins (Eglps), is thought to have caused the loss of the classical aquaglyceroporins in holometabolous and hemipteran insects. By contrast, with the exception of green algae (Chlorophyta), land plants (Embryophyta), and some single-celled organisms, glycerol transport is achieved via the classical aquaglyceroporins (Glps) in all other taxa studied to date. The neofunctionalization and gene replacement due to increased efficiency of glycerol transport represents a form of Darwinian evolution at the molecular level and can thus be considered as molecular supplantation.

Amongst protists and multicellular organisms, the aquaporin gene family is most diverse in algae, plants, and vertebrates. These lineages retain high gene copy numbers associated with polyploidy with up to 71 paralogs in upland cotton and at least 42 paralogs in Atlantic salmon (Fig. 3c). In green algae, eight types of aquaporin have been identified including plasma membrane intrinsic proteins (PIPs), algal-specific forms (MIPA-E), GlpF-like intrinsic proteins (GIPs), and small basic intrinsic proteins (SIPs), while the embryophyte superfamily consists of seven major phylogenetic groups, which evolved by both horizontal and vertical gene transfer (Anderberg et al. 2011; Abascal et al. 2014). The NIPs and GIPs, which represent the major glycerol transporters in plants, have their origins in bacteria, while PIPs, X intrinsic proteins (XIPs), and SIPs likely evolved in unicellular eukaryotes and were vertically transferred to algae and plants. Conversely, tonoplast intrinsic proteins (TIPs) and hybrid intrinsic proteins (HIPs) appear to have their origins in ancient ancestors of Embryophyta, which probably retained all seven of the aquaporin groups as exemplified by mosses (Danielson and Johanson 2008). Three of these gene groups, however, were subsequently lost prior to and during the evolution of flowering plants (Angiosperms), with GIPs and HIPs disappearing prior to the evolution of paired leaf-seed plants (Dicotyledonae) and XIPs prior to the evolution of single leaf-seed plants (Monocotyledonae) (Danielson and Johanson 2008).

Each of the four grades of aquaporin outlined above has been found in basal metazoan organisms, including Porifera (sponges) and Cnidaria (jellyfish and corals), although Aqp4-like (Aqp4L) and Aqp8-like (Aqp8L) channels have yet to be identified in Porifera and Cnidaria, respectively (Fig. 3c). These data nevertheless suggest that the four grades may have been an ancient feature of eukaryotic organisms, with the lack of a particular grade indicating lineage-specific gene loss. Amongst Protostomia, such gene loss is exemplified by Aqp4L channels in Nematoda and Aqp8L channels in Arthropoda (Finn and Cerdà 2015). All four grades are nevertheless present in basal Deuterostomia, with 17 subfamilies (Aqp0–Aqp16) now identified in vertebrates (Finn et al. 2014). Modern placental mammals, however, have only retained 13 functional forms of these subfamilies (Aqp0–Aqp12), while monotremes still harbor the Aqp13 and -14 paralogs. Selective tandem evolution of Aqp2, -5, and -6 paralogs, which play vital roles in the water conservation of all extant tetrapods, occurred exclusively in the sarcopterygian (lobed finned) lineage and is considered to have been permissive for the adaptation of vertebrates to land (Finn et al. 2014).

Karyotypic analyses show that the the aquaporin gene repertoires are distributed throughout the genomes of mammals, fish, and plants but show a more restricted distribution in insects (Fig. 4). Gene structures commonly consist of three exons for the unorthodox channels and between four and six exons for the orthodox channels, with overall gene length varying according to the taxonomic lineage. There is further a growing database of literature that documents mutliple splice variants of vertebrate and arthropod aquaporins, although in most cases, the functional significance of such isoforms remains to be established.
Aquaporin, Fig. 4

Genomic structures and karyotypic distributions of eukaryotic aquaporins. Annotated genes are shown to scale and localized in the genomes of human (Homo sapiens), zebrafish (Danio rerio), fruit fly (Drosophila melanogaster), and thale cress (Arabidopsis thaliana) based upon Ensembl. Solid boxes represent exons with grey triangles indicating coding direction. Gene clusters are linked by dotted lines. Numbers above broken introns indicate length in kilobases (kb)

Permeability

The molecular transport properties of aquaporins are typically tested experimentally using homologous or heterologous expression experiments in amphibian oocytes, cultured cell lines, and yeast cells or from reconstituted aquaporins in liposomes. Such experiments have revealed that nearly all aquaporins transport water, while certain grades including the Aqp8-related channels and aquaglyceroporins are multifunctional (Table 2). A number of exceptions to water transport have nevertheless been documented, including loss of this function in laboratory strains of yeast Aqy2, some plant PIP1s and NIP6s, and Drosophila Bib, which evolved into a cation transporter (K+ > Na+). Mammalian AQP0 has an apparent low-intrinsic water permeability; however, this is likely a derived feature, since piscine-Aqp0 orthologs, including both genomic duplicates in zebrafish and all four paralogs in tetraploid Atlantic salmon, transport water efficiently (Chauvigné et al. 2015). In addition to water, most eukaryotic channels are also considered capable of permeating hydrogen peroxide (H2O2) to mediate cell signaling or alleviate oxidative stress generated by mitochondria or chloroplasts (Bienert and Chaumont 2014). A number of vertebrate and plant aquaporins have been shown to transport gases, such as NH3 and CO2, although the molecular mechanism is still debated.

The multifunctional passage of larger solutes, such as glycerol, is an inherent property of the classical aquaglyceroporins, which typically retain small uncharged residues in P2 of the ar/R selectivity filter, rather than the conserved His of the classical water channels. Indeed, mutation of the His to Ala or Ser in the Eglps of hexapods converted these Aqp4-related channels into efficient glycerol transporters. Ion transport is rare for the aquaporin superfamily but has nevertheless evolved independently in the Drosophila Bib and mammalian AQP6 channels.

Regulation

Beyond osmotic pressure associated with changing solute concentrations, cellular control of the rate of molecular permeation through aquaporin pores is achieved primarily via gating while the channel is integrated in the target membrane or by trafficking to and from the membrane. Gating has been observed for plant, fungal, and vertebrate aquaporins where the orthologous pores are closed or opened by mechanical stress, altered [H+] and [Ca2+], or phosphorylation/dephosphorylation of specific residues. In each case, the creation or removal of additional electrochemical barriers or conformational changes is thought to mediate the open or closed state (Törnroth-Horsefield et al. 2010). For example, during drought stress, dephosphorylation of two conserved Ser residues in plant PIPs causes pore closure by extension of a Leu residue on loop D into the cytoplasmic region of the channel, but during flooding, reduced intracellular pH associated with anoxia also closes the pore through protonation of a conserved loop D His residue. The loop D Leu closure mechanism may also be invoked by the binding of Ca2+ ions, which causes hydrogen bond interactions between loop D, loop B, and N-terminal residues. In fungal Aqy1, an N-terminal Tyr residue is thought to be responsible for gating the cytoplasmic region of the pore, which can be modulated either through mechanical stress or potentially through phosphorylation/dephosphorylation of a loop B Ser residue. Mechanical stress in the form of pressure pulses is further suspected to invoke gating of some plant aquaporins. Conversely, extracellular pH and intracellular [Ca2+] regulate the channel permeability of vertebrate AQP0, where the position of a single His residue at the extracellular junction of transmembrane 2 and loop A is critical for pH-regulated water permeability, which varies between fish and mammals, while elevated intracellular [Ca2+] acting through calmodulin closes the pore. Extracellular acidity further converts AQP3 from a water and glyerol transporter to predominantly a glycerol transporter between pH 6.1 and 6.4 but closes the pore at pH < 6.1, while phosphorylation/dephosphorylation of a loop B Ser residue is suggested to regulate water permeation through AQP4, although this latter control remains controversial. With the exception of the gating role of plant PIPs, however, the physiological significance of channel gating still remains to be established.

The predominant regulatory mechanism of vertebrate aquaporins is trafficking. During membrane protein trafficking, protein sorting signals are recognized by components of the trafficking machinery, directing them to their target locations. For aquaporins, sorting signals involve multiple posttranslational modifications of the N- or C-termini, or the internal loops, typically involving phosphorylation/dephosphorylation events of specific residues. In mammals, the trafficking of AQP2 in the kidney collecting duct has been extensively investigated and represents the canonical example of aquaporin trafficking. This involves the antidiuretic hormone vasopressin that, upon dehydration, triggers the redistribution of AQP2 from intracellular storage vesicles to the apical membrane. During this mechanism, protein kinase A phosphorylation of Ser256 is a prerequisite for targeting to the apical membrane, whereas ubiquitination of Lys270 triggers AQP2 internalization, after which it can be stored for another round of vasopressin-mediated trafficking, or targeted for degradation. A number of proteins have also been shown to bind AQP2 and regulate its trafficking, predominantly via the AQP2 C-terminal region. Examples include heat shock protein 70 (hsp70/hsc70) and LYST interaction protein 5 (LIP5), as well as Spa-1. Protein-protein interactions involving the C-terminus have also been demonstrated to regulate AQP0, AQP4, and AQP5 trafficking (Sjöhamn and Hedfalk 2014).

Physiological Functions

The physiological roles of aquaporins have been best studied in mammals, where channel expression is widely distributed in the different tissues (Fig. 5). Loss-of-function, or aberrant expression has identified a number of maladies, such as AQP0-related cataract, AQP1-related diuresis, AQP2-related diabetes insipidus, AQP3-related skin tumors, AQP4-related hydrocephalus, AQP5-related Sjögren’s syndrome, AQP6-related Meniere’s disease, AQP7-related obesity, AQP8-related colitis, AQP9-related hyperglycemia, AQP10-related celiac disease, and AQP11-related polycystic kidneys. Extensive sources of recent information are available on the roles of mammalian aquaporins in the brain, in the kidney, in cancer biology, and in drug discovery (Verkman et al. 2014).
Aquaporin, Fig. 5

Aquaporin expression in humans. Expression patterns are annotated based upon the presence of mRNA or protein in a given tissue, with aquaporins colored according to their grade. Data are derived from the EMBL Expression Atlas, and literature sources cited in the text

In nonmammalian tetrapods, the water-balancing roles of aquaporins in the skin, urinary bladder, and mesonephric kidneys of anuran amphibians have been elucidated in a series of studies lead by Masakazu Suzuki (Suzuki et al. 2015). This includes the constitutive role of AQP3 in the basolateral membranes, the conserved role of AQP2 in the apical membranes of renal collecting ducts, the role of a urinary bladder type of AQP6 (AQP6ub) in the urinary bladder, and the role of two additional ventral skin-types of AQP6 (AQP6vs1 and -6vs2) in the ventral and thigh skin regions (Fig. 6a). The function of each channel is regulated by neurohypophysial peptides (arginine vasotocin and hydrins) to promote water uptake via the skin and recycling from the kidneys and urinary bladder under dehydrating conditions. Specific expression patterns are found in relation to the ecology of the species. For example, aquatic species do not express AQP6vs1 or –6vs2 channels in the ventral skin, while AQP6ub and AQP5 are found in the urinary bladder. Conversely, semiaquatic species express both AQP6vs1 and -6vs2 in the ventral and thigh skin, while arboreal and terrestrial species recruit the AQP6ub paralog to the ventral skin instead of AQP6vs2 (Suzuki et al. 2015) (Fig. 6b).
Aquaporin, Fig. 6

Aquaporin expression in anuran amphibians. (a) During dehydration, neurohypophysial peptides, arginine vasotocin (AVT), and hydrins induce water uptake via aquaporins in the ventral patch and thigh skin and recycling of water via aquaporins in the urinary bladder and kidneys. The figure is modified after Suzuki et al. (2015). (b) AQP2 functions as the major channel recycling water in the kidneys, but different paralogs of AQP6 (ventral skin-type or urinary bladder-type) and AQP5 are recruited for water recycling depending upon the ecology of the species

In fishes, the roles of aquaporins have been mostly studied in teleosts, where preliminary expression maps are compiled for zebrafish and other species (Fig. 7). The physiological significance of piscine aquaporins is best understood in the gametes. For example, water transport through a duplicated ortholog of mammalian AQP1, termed Aqp1ab, has been established in marine species that spawn highly hydrated eggs that float when spawned. Aqp1ab is temporarily expressed in the oocyte plasma membrane when osmotic pressure derived from proteolyzed yolk proteins is maximized to facilitate oocyte hydration (Cerdà et al. 2013). In marine teleost spermatozoa, a different duplicated ortholog of mammalian AQP8, termed Aqp8bb, functions as a peroxiporin. Upon seawater activation, Aqp8bb is recruited to the inner mitochondrial membrane to facilitate the efflux of hydrogen peroxide for the maintenance of membrane potential, ATP production, and the prolongation of spermatozoon motility (Chauvigné et al. 2015).
Aquaporin, Fig. 7

Aquaporin expression in teleosts. Colored dots refer to the tissue distribution in adult zebrafish, while white dots represent data reported for other species. In some species, the anterior intestine is indicated by the stomach, while the brain includes chemosensory and mechanosensory organs. The figure is modified after Finn and Cerdà (2011)

The roles of aquaporins in protostome animals have mainly focused on insects such as dipteran flies and mosquitoes, although some investigations have been conducted on nematodes, ticks, and sea lice. Preliminary expression maps indicate that the insect aquaporins, which consist of Bib, Drosophila intrinsic protein (Drip), Pyrocoelia rufa integral protein (Prip), Eglps, and the unorthodox Aqp12L channels, are widely distributed in the different tissues (Fig. 8). Physiological roles are thought to include Bib in cell-to-cell adhesion, cation transport and neurogenesis, Drip and Prip in oocyte hydration, lactation, saliva production, digestion and diuresis associated with blood or sap feeding, and adaptive desiccation, and Eglps in freeze tolerance, or alleviation from heat stress. A number of these roles are recently reviewed for blood-feeding arthropods (Benoit et al. 2014).
Aquaporin, Fig. 8

Aquaporin expression in insects. Colored dots refer to the tissue distribution in adult Drosophila, while white dots represent data reported for other species of insect. Data are compiled from the Fly Atlas and literature sources cited in the text. The image is modified under the GNU Free Documentation License

Amongst protists, the single channel or suite of paralogous aquaporins are important for solute transport, cell volume regulation in response to hyposmotic stress, osmotaxis, glycerol flux associated with energy metabolism, encapsulation, germination, and expulsion of toxic metabolic byproducts. For fungal yeasts, aquaporins and aquaglyceroporins are considered important for freeze tolerance, water transport during development, osmosensing, modulation of cell surface properties for substrate adhesion and formation of cell biofilms. Within filamentous fungi, aquaglyceroporins are considered important for fruiting body formation and in the water flow and nutrient exchange of ectomycorrhizas.

In land plants, the large array of aquaporin paralogs are not redundant but expressed in the roots, leaves, fruits, and flowers. The thale cress (Arabidopsis thaliana) is the best studied model organism in this regard with 33 of the 38 genomic paralogs expressed above 0.6 FPKM in these for tissues (Fig. 9a). PIPs and TIPS are by far the most prevalent types of channel expressed in each tissue. To facilitate the continuum of transpirated water from root to leaf, water exits the conducting xylem vessels and passes through the different cell layers either via the cell-to-cell or apoplastic pathways (Fig. 9b), while water is absorbed and traverses the root centripetally by the same pathways to hydrate the xylem vessels (Fig. 9c). In the apoplastic pathway, water does not enter the cell but passes through the space outside the plasma membrane (apoplast), while water entering and leaving via the cell-to-cell pathway may do so simplistically through cytoplasmic bridges (plasmodesmata) or transcellularly across the plasma membranes via aquaporins (Maurel et al. 2015). PIPs appear to be particularly important in this regard, while TIPs are typically expressed in the vacuolar membrane (tonoplast). PIPs and TIPs are also implicated in the detoxification and signaling of H2O2 between chloroplasts and vacuoles, in the hydralic control of growth, and in the CO2 conductance through mesopyllic cells. NIPs, which tend to be expressed at low levels in the roots and the membranes of rhizobia bacteroid-containing symbiosomes found in the root nodules of leguminous plants, are important for uptake of soil micronutrients. This includes boron, silicon, arsenic, antimony, selenium, and aluminium in the form of Al(OH)3, which are important for plant growth, health, or flower color (Maurel et al. 2015).
Aquaporin, Fig. 9

Aquaporin expression in plants. (a) Transcript expression levels of 33 aquaporin paralogs in the root, leaf, fruit, and flower of Arabidopsis thaliana. Transcript levels are shown as fragments per kilobase of exon per million fragments mapped (FPKM). Data are derived from the EMBL Expression Atlas under the Creative Commons Attribution 4.0 International License, and literature sources cited in the text. (b) Water transport pathways in leaves and (c) roots. Aquaporins are recruited to the cell-to-cell pathway for transcellular water flow. Redrawn from Maurel et al. (2015) with permission from The American Physiological Society

Summary

Despite eluding scientists for centuries, aquaporins are now found in virtually every organism studied, often with surprisingly broad repertoires in both prokaryotes and eukaryotes. The core transmembrane domains are conserved in all channels to form a central pore that primarily facilitates the passage of water and other small uncharged solutes, while the extracellular and intracellular loops and terminal domains are the major sites for regulation. Control of water and solute flux is typically achieved via the osmotic pressure, or gating and trafficking associated with altered pH, divalent ions, and phosphorylation/dephosphorylation events. The channels can play important homeostatic roles within intracellular compartments or between the plasma membrane and extracellular space, while a combination of different paralogs located in the apical or basolateral membranes of multicellular organisms is typically responsible for transcellular fluid transport. Although advances have been made in our understanding of aquaporin biology, there remains a general lack of crystallographic models for most of the identified orthologs. To date, there are no crystals available for metazoan aquaglyceroporins, Aqp8-type channels, or unorthodox aquaporins. Such data are essential for the accurate assessment of the structural mechanisms underlying the differences in water and solute transport, as well as their inhibition and molecular regulation, and will be of paramount importance for drug discovery. In this respect, the development of specific blockers will be highly advantageous for biomedical research. Very little is known of the function of the unorthodox channels, despite their broad expression patterns in insects and vertebrates. Similarly, although evolutionary studies have opened a window on the great variety of channels in prokaryotic and eukaryotic organisms, the positive selection and apparent nonredundant functions of multiple closely related channels have yet to be elucidated. Based upon the established biological importance of aquaporins in animal health, cancer, crop biology, and infective or parasitic diseases, however, the water channel proteins of different organisms remain highly attractive as therapeutic targets.

References

  1. Abascal F, Irisarri I, Zardoya R. Diversity and evolution of membrane intrinsic proteins. Biochim Biophys Acta. 2014;1840:1468–81.CrossRefPubMedGoogle Scholar
  2. Agre P, Sasaki S, Chrispeels MJ. Aquaporins: a family of water channel proteins. Am J Phys. 1993;265:F461.Google Scholar
  3. Anderberg HI, Danielsen JÅH, Johanson U. Algal MIPs, high diversity and conserved motifs. BMC Evol Biol. 2011;11:110.PubMedCentralCrossRefPubMedGoogle Scholar
  4. Benga G, Popescu O, Pop VI, Holmes RE. p-(Ch1oromercuri) benzenesulfonate binding by membrane proteins and the inhibition of water transport in human erythrocytes. Biochemistry. 1986;25:1535–8.CrossRefPubMedGoogle Scholar
  5. Benoit JB, Hansen IA, Szuter EM, Drake LL, Burnett DL, Attardo GM. Emerging roles of aquaporins in relation to the physiology of blood-feeding arthropods. J Comp Physiol. 2014;184B:811–25.CrossRefGoogle Scholar
  6. Bienert GP, Chaumont F. Aquaporin-facilitated transmembrane diffusion of hydrogen peroxide. Biochim Biophys Acta. 2014;1840:1596–604.CrossRefPubMedGoogle Scholar
  7. Cerdà J, Zapater C, Chauvigné F, Finn RN. Water homeostasis in the fish oocyte: New insights into the role and molecular regulation of a teleost-specific aquaporin. Fish Physiol Biochem. 2013;39: 19–27.CrossRefPubMedGoogle Scholar
  8. Chauvigné F, Boj M, Finn RN, Cerdà J. Mitochondrial aquaporin-8-mediated hydrogen peroxide transport is essential for teleost spermatozoon motility. Sci Rep. 2015;5:7789.PubMedCentralCrossRefPubMedGoogle Scholar
  9. Danielson JÅH, Johanson U. Unexpected complexity of the aquaporin gene family in the moss Physcomitrella patens. BMC Plant Biol. 2008;8:45.PubMedCentralCrossRefPubMedGoogle Scholar
  10. Denker BM, Smith BL, Kuhajda FP, Agre P. Identification, purification, and partial characterization of a novel Mr 28,000 integral membrane protein from erythrocytes and renal tubules. J Biol Chem. 1988;263:15634–42.PubMedGoogle Scholar
  11. Finn RN, Cerdà J. Aquaporin evolution in fishes. Front Physiol. 2011;2:44.PubMedCentralCrossRefPubMedGoogle Scholar
  12. Finn RN, Cerdà J. Evolution and functional diversity of aquaporins. Biol Bull. 2015;229:6–23.CrossRefPubMedGoogle Scholar
  13. Finn RN, Chauvigné F, Hlidberg JB, Cutler C, Cerdà J. The lineage-specific evolution of aquaporin gene clusters facilitated tetrapod terrestrial adaptation. PLoS One. 2014;9(11):e113686.PubMedCentralCrossRefPubMedGoogle Scholar
  14. Finn RN, Chauvigné F, Stavang JA, Belles X, Cerdà J. Insect glycerol transporters evolved by functional co-option and gene replacement. Nature Commun. 2015;6:7814.CrossRefGoogle Scholar
  15. Gorin MB, Yancey SB, Cline J, Revel JP, Horwitz J. The major intrinsic protein (MIP) of the bovine lens fiber membrane: characterization and structure based on cDNA cloning. Cell. 1984;39:49–59.CrossRefPubMedGoogle Scholar
  16. Heller H. The effect of neurohypophysial extracts on the water balance of lower vertebrates. Biol Rev. 1945;20:147–58.CrossRefPubMedGoogle Scholar
  17. Hevesy GV, Hofer V, Krogh A. The permeability of the skin of frogs to water as determined by D20 and H2O. Skand Arch Physiol. 1935;72:199–214.CrossRefGoogle Scholar
  18. Koefoed-Johnsen V, Ussing HH. The contributions of diffusion and flow to the passage of D2O through living membranes. Acta Physiol Scand. 1953;28:60–76.CrossRefPubMedGoogle Scholar
  19. Macey RL, Farmer REL. Inhibition of water and solute permeability in human red cells. Biochim Biophys Acta. 1970;211:104–6.CrossRefPubMedGoogle Scholar
  20. Maurel C, Boursiac Y, Luu DT, Santoni V, Shazad Z, Verdoucq L. Aquaporins in plants. Physiol Rev. 2015;95:1321–58.CrossRefPubMedGoogle Scholar
  21. Paganelli CV, Solomon AK. The rate of exchange of tritriated water across the human red cell membrane. J Gen Physiol. 1957;41:259–77.PubMedCentralCrossRefPubMedGoogle Scholar
  22. Pao GM, Wu LF, Johnson KD, Höfte H, Chrispeels MJ, Sweet G, Sandal NN, Saier Jr MH. Evolution of the MIP family of integral membrane transport proteins. Mol Microbiol. 1991;5:33–7.CrossRefPubMedGoogle Scholar
  23. Preston GM, Carroll TP, Guggino WB, Agre P. Appearance of water channels in Xenopus oocytes expressing red cell CHIP28 protein. Science. 1992;256:385–7.CrossRefPubMedGoogle Scholar
  24. Sjöhamn J, Hedfalk K. Unraveling aquaporin interaction partners. Biochim Biophys Acta. 2014;1840:1614–23.CrossRefPubMedGoogle Scholar
  25. Suzuki M, Shibata Y, Ogushi Y, Okada R. Molecular machinery for vasotocin-dependent transepithelial water movement in amphibians: aquaporins and evolution. Biol Bull. 2015;229:109–19.CrossRefPubMedGoogle Scholar
  26. Törnroth-Horsefield S, Hedfalk K, Fisher G, Lindkvist-Petersson K, Neutze R. Structural mechanism of plant aquaporin gating. FEBS Lett. 2010;584:2580–8.CrossRefPubMedGoogle Scholar
  27. Verkman AS, Anderson MO, Papadopoulis MC. Aquaporins: important but elusive drug targets. Nat Rev Drug Discov. 2014;13:259–77.PubMedCentralCrossRefPubMedGoogle Scholar
  28. Wspalz T, Fujiyoshi Y, Engel A. The AQP structure and functional implications. In: Beitz E, (ed). Aquaporins. Handb Exp Pharmacol. 2009; 190:31–56.Google Scholar

Copyright information

© Springer International Publishing AG 2018

Authors and Affiliations

  1. 1.Department of Biology, Bergen High Technology CentreUniversity of BergenBergenNorway
  2. 2.Institute of Marine ResearchBergenNorway
  3. 3.Institut de Recerca i Tecnologia Agroalimentàries (IRTA)Consejo Superior de Investigaciones Científicas (CSIC)BarcelonaSpain
  4. 4.Institut de Recerca i Tecnologia Agroalimentàries (IRTA)-Institut de Ciéncies del MarConsejo Superior de Investigaciones Científicas (CSIC)BarcelonaSpain