14. Bergmeyer, H. U. Methods of Enzymatic Analysis (Verlag Chemie, Weinheim/Bergstr., 1974).
15. Kushnir, M. M., Komaromy-Hiller, G., Shushan, B.,Urry, F. M. & Roberts, W. L. Analysis of
dicarboxylic acids by tandem mass spectrometry. High-throughput quantitative measurement of
methylmalonic acid in serum, plasma, and urine. Clin. Chem. 47, 1993–2002 (2001).
16. Martin, M., Ferrier, B. & Baverel, G. Transport and utilization of alpha-ketoglutarate by the rat kidney
in vivo. Pflugers Arch. 413, 217–224 (1989).
17. Gullans, S. R., Kone, B. C., Avison, M. J. & Giebisch, G. Succinate alters respiration, membrane
potential, and intracellular K
þ
in proximal tubule. Am. J. Physiol. 255, F1170–F1177 (1988).
18. Gullans, S. R., Brazy, P. C., Dennis, V. W. & Mandel,L. J. Interactions between gluconeogenesisand
sodium transport in rabbit proximal tubule. Am. J. Physiol. 246, F859–F869 (1984).
19. Krebs, H. A. Rate control of the tricarboxylic acid cycle. Adv. Enzyme Regul. 8, 335–353 (1970).
20. Hems, D. A. & Brosnan, J. T. Effects of ischaemia on content of metabolites in rat liver and kidney in
vivo. Biochem. J. 120, 105–111 (1970).
21. Pan, L. et al. Critical roles of a cyclic AMP responsive element and an E-box in regulation of mouse
renin gene expression. J. Biol. Chem. 276, 45530–45538 (2001).
22. Hackenthal, E., Paul, M., Ganten, D. & Taugner, R. Morphology, physiology, and molecular biology of
renin secretion. Physiol. Rev. 70, 1067–1116 (1990).
23. An, S. et al. Identification and characterization of a melanin-concentrating hormone receptor. Proc.
Natl Acad. Sci. USA 98, 7576–7581 (2001).
24. Brandish, P. E., Hill, L. A., Zheng, W. & Scolnick, E. M. Scintillation proximity assayof inositol
phosphates in cell extracts: high-throughput measurement of G-protein-coupled receptor activation.
Anal. Biochem. 313, 311–318 (2003).
25. Chuang, P. T., Kawcak, T. & McMahon, A. P. Feedback control of mammalian Hedgehog signaling by
the Hedgehog-binding protein, Hip1, modulates Fgf signaling during branching morphogenesis of
the lung. Genes Dev. 17, 342–347 (2003).
26. Palczewski, K. et al. Crystal structure of rhodopsin: A G protein-coupled receptor. Science 289,
739–745 (2000).
27. Krege, J. H., Hodgin, J. B., Hagaman, J. R. & Smithies, O. A noninvasive computerized tail-cuff system
for measuring blood pressure in mice. Hypertension 25, 1111–1115 (1995).
28. Sugiyama, F. et al. QTL associated with blood pressure, heart rate, and heart weight in CBA/CaJ and
BALB/cJ mice. Physiol. Genomics 10, 5–12 (2002).
29. Silva, A. P. et al. Bilateral nephrectomy delays gastric emptying of a liquid meal in awake rats. Ren. Fail.
24, 275–284 (2002).
30. Woronicz,J. D., Gao, X., Cao, Z., Rothe,M. & Goeddel, D. V. IkB kinase-b: NF-kB activation and
complex formation with IkB kinase-a and NIK. Science 278, 866–869 (1997).
Supplementary Information accompanies the paper on www.nature.com/nature.
Acknowledgements We thank J. Reagan for critical comments during ligand identification;
L. Yang for sharing reagents for in situ hybridization; G. Cutler, J. Knop, H. Baribault, J. Ma,
S.-C. Miao, W. Inman, C. Ogden, S. Shuttleworth and M. Rich for providing support and
discussions; and D. Goeddel, B. Lemmon and T. Hoey for critical reading of the manuscript.
Competing interests statement The authors declare that they have no competing financial
interests.
Correspondence and requests for materials should be addressed to L.L. (ling@tularik.com).
..............................................................
Aquaporin-0 membrane junctions
reveal the structure of a closed
water pore
Tamir Gonen
1
, Piotr Sliz
2
, Joerg Kistler
3
, Yifan Cheng
1
& Thomas Walz
1
1
Department of Cell Biology, Harvard Medical School, 240 Longwood Avenue,
Boston, Massachusetts 02115, USA
2
Howard Hughes Medical Institute and Children’s Hospital Laboratory of
Molecular Medicine, 320 Longwood Avenue, and Department of Biological
Chemistry and Molecular Pharmacology, Harvard Medical School, 240 Longwood
Avenue, Boston, Massachusetts 02115, USA
3
School of Biological Sciences, University of Auckland, Auckland PO Box 92019,
New Zealand
.............................................................................................................................................................................
The lens-specific water pore aquaporin-0 (AQP0) is the only
aquaporin known to form membrane junctions in vivo
1
. We show
here that AQP0 from the lens core, containing some carboxy-
terminally cleaved AQP0
2,3
, forms double-layered crystals that
recapitulate in vivo junctions. We present the structure of the
AQP0 membrane junction as determined by electron crystal-
lography. The junction is formed by three localized interactions
between AQP0 molecules in adjoining membranes, mainly
mediated by proline residues conserved in AQP0s from different
species but not present in most other aquaporins. Whereas all
previously determined aquaporin structures show the pore in an
open conformation
4–9
, the water pore is closed in AQP0 junc-
tions. The water pathway in AQP0 also contains an additional
pore constriction, not seen in other known aquaporin struc-
tures
4–9
, which may be responsible for pore gating.
AQP0 is a member of the aquaporin family, members of which
form pores that are either highly selective for water (aquaporins) or
also permeable to other small neutral solutes such as glycerol
(aquaglyceroporins) (reviewed in ref. 10). To date, the atomic
structures of three aquaporins have been determined (AQP1
4–6
,
GlpF
7,8
and AQPZ
9
). Sequence alignment shows AQP0 to be closely
related to the pure water channel AQP1 (43.6% identity, 62.6%
similarity). The presence of His 172, a residue conserved only in
aquaporins but substituted in aquaglyceroporins, also suggests that
AQP0 forms a pure water pore. AQP0 water permeability at neutral
pH is approximately 40 times lower than that of AQP1
11
, but AQP0
water conductance doubles under mildly acidic conditions
12
. In the
case of aquaporins in plant roots, a pH-dependent closure of the
water pores has been reported
13
. Thus, evidence suggests that
certain aquaporin pores are gated.
AQP0 water pores are considered essential for the lens micro-
circulation system, proposed to supply deeper-lying fibre cells with
nutrients and to clear waste products
14,15
. Unlike all other aqua-
porins, AQP0 is also present in membrane junctions. It is particu-
larly enriched in the 11–13 nm thin junctions between lens fibre
cells, that feature square AQP0 arrays
1
. Atomic force microscopy
analysis of in vitro reconstituted AQP0 two-dimensional crystals
demonstrated these crystals to be double-layered
16
.
Using AQP0 from the core of sheep lenses, where some of the
AQP0 is proteolytically cleaved near the C terminus at various sites
in an age-dependent manner
2,3
, we reproduced the double-layered
two-dimensional crystals
16,17
. When core AQP0 was reconstituted at
a lipid-to-protein ratio of 0.25 (w/w), large membrane sheets
formed (.6 mm) that in some cases showed two parallel edges,
revealing them to be double-layered (Fig. 1a). The crystals showing
p422 symmetry had lattice constants of a ¼ b ¼ 65.5 A
˚
and a
thickness of 11 nm (Fig. 2a), the same dimensions as thin junctions
in the lens
1
. Double-layered AQP0 two-dimensional crystals are
therefore likely to recapitulate thin lens fibre cell junctions.
Electron diffraction analysis of AQP0 crystals (tilted to an angle
of up to 708) produced strong diffraction spots to 3 A
˚
resolution in
all directions (Fig. 1b, c; the electron crystallographic data are
summarized in Table 1). As the crystal structure of the homologous
bovine AQP1 was available
5
, we determined the structure of the
AQP0 membrane junction by molecular replacement, thus avoiding
the cumbersome and time-consuming process of collecting high-
resolution images of tilted specimens. Sequencing of cloned sheep
AQP0 showed an identical amino acid sequence to bovine AQP0
18
,
with the exception of three conservative (S20T, M90V and S240T)
and one non-conservative substitutions (C14F).
Our model (Fig. 2a) shows unique features that enable AQP0 to
form membrane junction interactions. These differ from those
previously suggested based on atomic force microscopy data
16
.
The extracellular surface of AQP0 is rather flat and the interactions
are mediated by direct contacts of the corresponding loops in the
opposing AQP0 molecules (Fig. 2a). Loop C, connecting a-helices
three and four, is significantly shorter than in AQP1 and GlpF. The
shortened loop C (also seen in AQP2, AQP5, AQP6 and AQP8) is
crucial for the formation of the very tight AQP0 junction, as it
allows three specific interactions to be formed that are mediated
almost exclusively by proline residues.
The most striking interaction involves Pro 38, in extracellular
loop A. The proline residues (Pro 38) from eight symmetry-related
AQP0 molecules in the stacked tetramers come together to form a
letters to nature
NATURE | VOL 429 | 13 MAY 2004 | www.nature.com/nature 193
©2004 Nature Publishing Group