| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Zentrum Physiologie, Abt. Vegetative Physiologie, Medizinische Hochschule Hannover, Hannover, Germany;
* Department of Cellular and Molecular Physiology, Yale School of Medicine, New Haven, Connecticut, USA;
Dipartimento di Chimica, Laboratorio di Chimica Bioinorganica, University of Florence, Sesto Fiorentino, Firenze, Italy; and
Department of Medicine, The Johns Hopkins University, Baltimore, Maryland, USA
3Correspondence: Zentrum Physiologie, Abt. Vegetative Physiologie, Medizinische Hochschule Hannover, 30625 Hannover, Germany. E-mail: gros.gerolf{at}mh-hannover.de
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: human red cell membrane • CO2 permeability • pCMBS • DIDS
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. The second was the observation that the water channel aquaporin-1 (AQP1) appears to serve as a conduit for CO2 (2
3
4)
, at least in artificial systems. The third was the finding that DIDS markedly reduces CO2 permeability of the red blood cell (RBC) membrane (5
, 6)
. The first evidence for the physiological relevance of gas channels was the demonstration that AQP1 plays a key role in CO2 uptake during photosynthesis by tobacco plants (7)
.
To determine the CO2 permeability of red cells and oocytes, we used a previously developed method of quantitating the CO2 permeability of cells in suspension by the mass spectrometric 18O exchange technique (8
9
10)
, which is especially suitable to detect rather high CO2 permeabilities that are difficult to measure by stopped-flow techniques. We also developed a new technique to visualize increases in CO2 permeation across membranes by measuring alkaline surface pH transients associated with augmented transfer of molecular CO2 from the extra- to the intracellular space of Xenopus laevis oocytes. From experiments in which we compare red blood cells from normal and Colton null (AQP1–/–) humans, we report that AQP1 appears to be responsible for 60% of the CO2 permeability, a finding confirmed by studies of CO2 permeability and alkaline surface pH transients in AQP1-expressing oocytes.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, 12)
. All red cells were washed three times in physiological saline before being used in the mass spectrometric experiments and were controlled for hemolysis. Hematocrit and blood cell count were taken to determine mean corpuscular volumes. Informed consent was sought and given in accordance with the Declaration of Helsinki. Blood samples were shipped chilled and red cells were used within 2–4 days after the samples were taken. It was ascertained by control experiment that this prolonged time interval between blood withdrawal and determination of permeability does not affect CO2 permeability values (PCO2): red cells from normal blood investigated on the day of withdrawal showed PCO2 of 0.16 ± 0.06 cm/s (n=35), while another group of normal red cell samples, after having traveled for several days, yielded PCO2 of 0.16 ± 0.07 cm/s (n=20).
Inhibitors
Most mass spectrometric experiments with intact red cells were carried out in the presence of the extracellular carbonic anhydrase (CA) sulfonamide inhibitor 1-[5-sulfamoyl-1,3,4-thiadiazol-2-yl-(aminosulfonyl-4-phenyl)]-2,4,6-trimethyl-pyridinium perchlorate, STAPTPP, described by Casey et al. (13)
, at a final concentration of 1 · 10–5 M. The inhibitor was added in order to ensure inhibition of small amounts of CA that may be set free by hemolysis during the mass spectrometric experiment due to the continuous stirring of the red cell suspension. We demonstrated that the intracellular CA activity of red cells was not affected by 30 min preincubation of red cells with the same concentration of this inhibitor. Para-(chloromercuri)-benzenesulfonate, or pCMBS, is a known inhibitor of the water as well as CO2 permeability of AQP1 (3
, 12
, 14)
. 4,4'-Diisothiocyanato-2,2'-stilbenedisulfonate (DIDS) was used as an established inhibitor of the red cell Cl–-HCO3– exchanger AE1 (band 3) and other ion transporters, which we have previously shown to also inhibit red cell CO2 permeability (5)
.
AQP1 expression in Xenopus laevis oocytes
Oocyte isolation was performed according to established procedures (15)
. Briefly, female Xenopus laevis were anesthetized using 0.2% MS-222 (ethyl 3-aminobenzoate methanesulfonate; Sigma, St. Louis, MO, USA). Ovarian lobes were removed and placed in 0 Ca solution (98 mM NaCl, 2 mM KCl, 1 mM MgCl2, 5 mM HEPES, pH 7.5). Enzymatic dissociation and defolliculation of oocytes were performed using 2 mg/ml type IA collagenase (Sigma) in 0 Ca solution. Stage V-VI oocytes were selected and kept at 18°C in sterile-filtered OR3 medium containing (per 2 L) one pack of powdered Leibovitz L-15 media with L-glutamine (GIBCO-Life Technologies, Inc., Gaithersburg, MD, USA), 100 ml of 10,000 U penicillin, 10,000 U streptomycin solution (Sigma), and 5 mM HEPES, pH 7.5.
Capped cRNA was transcribed from a plasmid containing human AQP1 (kind gift from Peter Agre, Johns Hopkins University) or mouse CA II (kind gift from William S. Sly, Saint Louis University) using Message Machine kit (Ambion, Austin, TX, USA). Oocytes were injected with 50 nl 0.02 µg/µl cRNA encoding CAII or with a mixture containing cRNAs encoding CAII (0.02 µg/µl) and AQP1 (0.005 µg/µl). Experiments were performed 3–6 days after injection. Expression of AQP1 in oocytes was substantiated by determining the time from their immersion in deionized water to lysis (2)
. This time was >20 min in control oocytes and <1 min in AQP1-expressing oocytes. Similarly, the expression of CA was verified by determining the oocytes’ CA activity by mass spectrometry, yielding an activity of 
10 U for control oocytes and of 150–200 U for CAII-expressing oocytes.
Principle of mass spectrometric measurement
We adapted the technique of Itada and Forster (16)
to study the exchange of 18O-labeled CO2 and HCO3– between an extracellular solution and the intracellular space of cells containing CA activity, in the present study either red cells in suspension or oocytes expressing CA also suspended in fluid. Itada and Foster (16)
showed that by observing the time course of the species C18O16O (mass/charge, m/z, 46 peak height) in the suspension with a mass spectrometer, it is possible to determine the intracellular CA activity Ai and the bicarbonate permeability PHCO3– of the cell membranes. The new feature of the present approach is the theoretical treatment of the process of 18O exchange, which does not assume, as Itada and Forster had, that CO2 is infinitely permeable and cannot be a limiting step in the overall exchange process (8
, 10)
.
In normal human red cells, CO2 permeation across the cell membrane has a detectable effect on the kinetics of 18O exchange and thus it is possible to determine the membrane CO2 permeability from measurements of 18O exchange kinetics (8
, 5)
. The set of equations describing the entire process of membrane transport of 18O-labeled CO2, HCO3–, and H2O as well as the chemical reactions between them has been reported (8
, 10)
, and is used here to estimate PCO2 from the kinetics of the extracellular concentration of C18O16O in suspensions of cells. These equations were solved numerically and a fitting procedure was used to find best-fit values of membrane PCO2 and PHCO3–. The value of the intracellular CA activity Ai, which is required for the calculations, was determined by measuring by mass spectrometry the CA activity of red cell lysates under conditions of pH and [Cl–] as they prevail inside cells.
Surface pH and Pf measurements in oocytes
Others (17
, 18)
had monitored transient changes in extracellular pH caused by the addition of CO2/HCO–3 to the extracellular fluid. We monitored the surface pH (pHS) of Xenopus oocytes—injected either with 50 nl of H2O or 50 nl of a solution containing RNA (0.5 ng/nl) encoding human AQP1—using a glass microelectrode (tip diameter of 15 µm), containing a pH-sensitive ionophore (Hydrogen Ionophore I-Cocktail B #95293, Fluka). When pHS was relatively stable, we periodically calibrated the pH electrode by using a model MPC-200 motorized micromanipulator (Sutter Instrument Co., Novato, CA, USA) to move the tip 300 µm away from the oocyte surface. At other times, the electrode tip pushed up against the oocyte surface 60 µm. The pH signal was amplified by a model FD223 high-impedance electrometer (World Precision Instruments, Inc., Sarasota, FL, USA). The external reference electrode was a calomel half cell with a 3-M KCl bridge. The CO2/HCO–3-free extracellular ND96 solution contained (in mM): 96 NaCl, 2 KCl, 1 MgCl2, 1.8 CaCl2, 5 HEPES, was titrated to pH 7.50, and had an osmolality of 200 mosmol/kg H2O. The CO2/HCO–3 solution contained (in mM): 66 NaCl, 2 KCl, 1 MgCl2, 1.8 CaCl2, 5 HEPES, 33 NaHCO3.
We measured water permeabilities, Pf, according to the method of Preston et al. (19)
. As described by Virkki et al. (20)
, we used a video camera to monitor the projection area of an oocyte while exposing the cell to a solution similar to CO2/HCO–3-free solution, except that we reduced [NaCl] to lower the osmolality to 80 mOsm/kg. A small metal sphere next to the oocyte served as a size reference. All experiments were performed at room temperature (22°C).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
1 cm/s. The same group of 85 measurements yielded a PHCO3– of 1.3 · 10–3 ± 0.3 · 10–3 (SD) cm/s, with a median of 1.3 · 10–3 cm/s.
Effect of pCMBS on CO2 permeability
Figure 1
illustrates an experimental series studying the effects of various concentrations of pCMBS on PHCO3– and PCO2. pCMBS is an efficient inhibitor of the water permeability of AQP1 (14)
, but is also known to inhibit band 3 protein to some extent (ID50 for pCMBS binding to band 3 was reported to be 2.0 mM by Zhang and Solomon, ref. (21)
. If PCO2 of Colton null red cells, 0.06 cm/s (Fig. 2
), is taken as the value to be expected with full inhibition of the AQP1 CO2 pathway, then the data of Fig. 1
lead to an estimate of I50 for inhibition of this pathway of 0.5 mM pCMBS. This is somewhat higher than the KI estimated for inhibition of the AQP1 water pathway by pCMBS of 0.13 mM (14)
, but is markedly lower than the I50 of pCMBS for band 3. In conclusion, pCMBS inhibits the CO2 permeability of human red cells at concentrations of 4-fold lower than needed for inhibition of band 3 and it will be shown below that this occurs by the action of pCMBS on AQP1.
|
|
CO2 permeability of aquaporin 1-deficient red cells
When Colton null (AQP1–/–) red cells are compared to normal human red cells, PHCO3– is identical in both types of red cells (Fig. 2A
) but PCO2 is significantly (by 60%) lower in Colton null cells than in control red cells (Fig. 2B
), indicating that AQP1 contributes significantly to CO2 permeation across the red cell membrane. Consistent with this, the AQP1 inhibitor pCMBS at concentrations of 1 and 2 mM has no significant effects on PCO2 of AQP1-deficient red cells, but it significantly reduces PCO2 of control cells (Fig. 2B
). Even at the higher pCMBS concentration of 3 mM, where the inhibitory effect of pCMBS on band 3 becomes apparent by a significant reduction of PHCO3– to 1/2 both in control and Colton null red cells (Fig. 2A
), this drug still produces no effect on PCO2 in Colton null red cells (Fig. 2B
). We conclude that pCMBS at all concentrations studied is without effect on PCO2 of AQP1-deficient red cells, confirming that pCMBS reduces PCO2 solely by its action on AQP1.
We have shown that the AE1 inhibitor DIDS markedly reduces PCO2 as well as PHCO3– in normal red cells (5)
. DIDS reduces PHCO3– to a similar extent in Colton null and in control cells (Fig. 2A
), as expected. On the other hand, PCO2, which is reduced in Colton null cells even in the absence of DIDS, is further reduced by DIDS to 0.012 cm/s, 1/5th of the figure obtained in the absence of DIDS. The lowest PCO2 value seen in Fig. 2B
, the value for Colton null cells treated with DIDS (0.012 cm/s), amounts to only 1/12th of the PCO2 found in normal red cells. In conclusion, PCO2 of AQP1-deficient red cells is reduced compared to control cells; it is further reduced by DIDS but is not affected by pCMBS. The inhibitory effect of DIDS lowers PCO2 in normal red cells by 0.10–0.11 cm/s, but lowers PCO2 of Colton null red cells by only 0.04–0.03 cm/s. These data suggest that part of the inhibitory effect of DIDS occurs by an action on AQP1, whereas the other part results from its action on a different DIDS-sensitive pathway.
Surface pH transients of AQP1-expressing oocytes and effect of DIDS
To test the hypothesis that DIDS reduces the CO2 permeability of AQP1, we examined the effect of DIDS in oocytes either expressing human AQP1 or injected with H2O. Preliminary experiments (not shown) indicated that 100 µM DIDS reduces the rate at which the addition of CO2 causes intracellular pH to fall in oocytes with vitelline membrane removed. Figure 3
A, B illustrates more sensitive experiments on oocytes (with vitelline membrane intact), in which we monitored the transient rise in extracellular surface pH caused by switching to an extracellular solution containing 5% CO2/33 mM HCO3– (pH 7.50). Adding CO2/HCO3– causes a rapid rise in pHS, followed by an approximately exponential decay. This pHS trajectory reflects the effect of CO2 influx (less any effect of HCO3– influx) on the equilibrium CO2 + H2O 
HCO3– + H+ near the cell surface. The spike height, an index of maximal CO2 influx, was substantially less in the H2O-injected oocyte (Fig. 3A
) than in the AQP1-expressing oocyte (Fig. 3B
), confirming that AQP1 increases CO2 permeability. After removal of CO2/HCO3–, we applied 100 µM DIDS and—in the continuous presence of DIDS—exposed each oocyte to CO2/HCO3– a second time. DIDS had little effect on spike height in the H2O-injected oocyte but reduced spike height in the AQP1-expressing cell.
|
Figure 3C
summarizes similar paired data from a total of six experiments each on H2O-injected and AQP1-expressing oocytes, and indicates that expression of AQP1 increases spike height by 96% and that DIDS reduces the AQP1-dependent spike height by 49%. Figure 3D
summarizes experiments in which we found that DIDS has no effect on Pf. Thus, DIDS substantially reduces the CO2 permeability of AQP1 without affecting H2O permeability.
Effects of aquaporin 1 expression on CO2 permeability of Xenopus laevis oocytes
Figure 4
gives the results of PCO2 determinations obtained with oocytes in suspension. In oocytes expressing CA II only, we measured a PCO2 of 0.057 cm/s, while coexpressing both CA and AQP1 raised PCO2 significantly to 0.11 cm/s. This agrees with the finding of Fig. 3
of a marked increase in the alkaline surface pH transient due to the expression of AQP1 in oocytes. The increase of PCO2 in AQP1-expressing oocytes complements the finding of a reduced PCO2 in aquaporin-1-deficient red cells.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
that the present technique is highly insensitive toward unstirred layers, likely because there is a store of bicarbonate in the external solution, including the membrane surface, which can replenish CO2 that has permeated the membrane into the cell interior. As the former assumption is likely not to be entirely fulfilled, the apparent PCO2 will tend to be an underestimate. It may be noted that from steady-state measurements of CO2 fluxes across layers of red cells, we earlier postulated a PCO2 of 1–2 cm/s (22
, 23)
.
The lipid phase of the membrane is likely to impart a "basal" CO2 permeability to the membrane, which may be represented by the value of 0.012 cm/s observed in AQP1-deficient red cells exposed to DIDS (Fig. 2B
). This latter value represents the lowest CO2 permeability observed in this study and would be 30-fold lower than the CO2 permeability of 0.3 cm/s observed by Gutknecht et al. (24)
in artificial lipid bilayers, whose composition, however, was not representative of the lipids in the red cell membrane and which were devoid of membrane proteins. Our data indicate that AQP1 and an additional DIDS-sensitive membrane protein markedly enhance red cell PCO2 to the point that the resistance of the membrane for CO2 diffusion is at least two orders of magnitude lower than that for HCO3–.
Bicarbonate permeabilities reported here are 1.5- to 3-fold higher than values reported in the literature, and their inhibition by DIDS appears less than has been established by other methods. This is due to some lysis that occurs by the stirring during the mass spectrometric experiment. It can be shown theoretically that the CA set free by this lysis will, before the inhibition by STAPTPP is complete, effect an apparent increase in PHCO3– of up to 4-fold, but will cause an overestimation of PCO2 by no more than a few percent.
Aquaporin-1-mediated permeation of CO2 across membranes
Aquaporin-1 constitutes a pathway for CO2
The absence of AQP1 from the red cell membrane reduces CO2 permeability from 0.15 cm/s to 0.06 cm/s (i.e., by 60%). The abundance of band 3, Glut1, the urea transporter, and spectrin are similar in AQP1-deficient and normal human red cells (12)
, supporting our conclusion that the decrease observed in PCO2 must be attributed to aquaporin-1. This conclusion is confirmed here by two other findings that were obtained by entirely different experimental approaches: 1) the 2-fold increase in PCO2 of the oocyte membrane on AQP1 expression as determined by mass spectrometry, and 2) the marked increase of the alkaline surface pH transient on AQP1-expressing oocytes. Our findings for PCO2 in oocytes are consistent with earlier studies examining CO2 influx into oocytes from the time course of intracellular pH (2
, 3)
.
There has been some controversy as to whether aquaporin-1 can act as a CO2 pore. Yang et al. (25)
observed the intraerythrocytic acidification on mixing of red cells with a CO2 solution in a rapid reaction, stopped-flow spectrophotometer and obtained similar PCO2 values for red cells from wild-type (WT) mice (0.012 cm/s) and for red cells from AQP1-deficient mice (0.011 cm/s). Similarly, they found identical PCO2 values of 10–3 cm/s in liposomes with and without reconstitution of AQP1 using stopped-flow. They concluded that aquaporin 1 does not act as a CO2 channel. Of note, PCO2 of 10–3 cm/s for liposomes is 100-fold lower that the value reported by Gutknecht et al. (24)
, and the PCO2 value observed for red cells is > 10-fold lower than the figures reported here. These differences could be due to considerable unstirred layers around the cells or, more likely, to inadequate mixing in the stopped-flow chamber that caused an apparently low CO2 permeability; therefore, the contribution of AQP1 was not evident. Our findings are in line with those of Prasad et al. (4)
, who used stopped-flow spectrophotometry to demonstrate a CO2 permeability in liposomes of 0.5 cm/s and in AQP1 reconstituted proteoliposomes of 1.9 cm/s, a nearly 4-fold increase that was inhibited by the aquaporin inhibitor HgCl2.
pCMBS inhibits the CO2 permeability via AQP1 only
The present data confirm previous reports that pCMBS inhibits a pathway for molecular CO2 across aquaporin 1-expressing oocyte membranes (3)
and across normal human red cell membranes (6)
. The data of the present paper show that in red cells this pathway is constituted by AQP1: pCMBS, an established inhibitor of the water permeability of AQP1 (14)
, at concentrations between 2 and 3 mM reduces PCO2 of normal red cells by 50–70% (Figs. 1
, 2)
but has no effect on PCO2 of Colton null red cells.
DIDS inhibits partially the CO2 pathway via AQP1
In normal human red cells DIDS reduces PCO2 from 0.15 cm/s to 0.05–0.04 cm/s, producing a 
PCO2 of 0.1 cm/s, whereas in Colton null red blood cells it reduces PCO2 from 0.06 cm/s to 0.012 cm/s, causing a PCO2 of 0.05 cm/s (Fig. 2B
). This is most simply interpreted to indicate that one-half of the DIDS effect is due to inhibition of AQP1 by DIDS, causing a PCO2 of 0.05 cm/s, and that the other half of the DIDS effect is due to an effect other than that on AQP1. This is most clearly substantiated by the inhibitory effect of DIDS on the surface pH transient observed in aquaporin-expressing oocytes (Fig. 3)
.
DIDS inhibits in addition a CO2 pathway that is not linked to AQP1
The effect of DIDS on PCO2 seen in AQP1-deficient red cells (Fig. 2B
) shows clearly there is another pathway for CO2 in the red cell membrane that 1) is inhibitable by DIDS, 2) is independent of AQP1, and 3) probably is not inhibited by pCMBS. Although it cannot be excluded that this pathway is constituted by the lipid bilayer and DIDS acts in an unknown manner on its lipid phase, as speculated by Forster et al. (5)
, it should be noted that DIDS has no effect in oocytes that do not express AQP1 (Fig. 3)
, and thus it appears more likely that this CO2 pathway is provided by another DIDS-inhibitable membrane protein of the red cell, such as AE1 (band 3), the Cl–-HCO3– exchanger. The identity of this protein has yet to be determined. From the numbers given above, its contribution may make up 32% of normal CO2 permeability, while the basal permeability due to the membrane lipids may be responsible for 8%.
Pathway of CO2 across the AQP1 molecule
What is the pathway in the aquaporin 1 homotetramer through which the CO2 molecule passes? It is well established that within the AQP1 tetramer each monomer possesses a water channel. Some have proposed that, in addition, AQP1 exhibits a nonselective cation conductance mediated by the single central pore of the homotetramer (26)
. Each AQP1 monomer possesses a curvilinear water channel, which exhibits wide openings on the extracellular and the cytoplasmic sides, and within the membrane narrows down to a size-selective pore in the center of the bilayer. This channel is 2.8–4 Å at its minimal diameter and has a length of 18–20 Å (27
28
29)
. Half of the channel surface is hydrophilic and the other half hydrophobic. Water with its size of 2.8 Å is thought to strip its waters of hydration via several water binding regions along the pore, thus allowing passage of a single water molecule through the pore. Ions are excluded on the basis of their charge and, in the hydrated form, their size (29)
. The channel does not conduct protons. The region around the 4-fold symmetry axis of the tetramer has been proposed to constitute the cation pore of aquaporin (26)
. It exhibits the narrowest constriction close to the external side of the pore; with an 3 Å diameter, its interior is largely hydrophobic. It possesses a large central cavity near the membrane midpoint and has many similarities with K+ channels (26
27
28)
. The cation channel appears to be cGMP-gated. Carbon dioxide with a MW of 44 is larger than water with its MW of 18, but has an effective molecular diameter between 2.8 and 3.4 Å, which is only slightly greater than the diameter of water of 2.8 Å (30
, 31)
.
Because CO2 is a linear molecule with a smaller transversal than longitudinal diameter, it may be expected to be able to permeate either of the two types of pores, each of which has a constriction site of 3 Å. CO2 is a quadrapole, that is, each of the two C-O bonds has a small charge separation. This may explain the observation that CO2, which is highly soluble in water and even more soluble in nonpolar solvents such as benzene, toluene, and heptane, exhibits its highest solubility in solvents such as acetone or ethanol, which contain both hydrophobic and strongly polar groups (32)
. In view of these properties, it is reasonable to propose that CO2 can transit the water channel of AQP1 with its mixed hydrophobic-hydrophilic surface. This conclusion would be in line with the observation by Cooper and Boron (3)
, who reported that pCMBS, though significantly reducing the rate of CO2 uptake and intracellular acidification rate in AQP1 expressing oocytes, does not affect CO2 permeation in the AQP1 mutant with serine replacing the mercury-sensitive cysteine (C189S) (33)
. Since C189 is close to the inner surface of the water channel of the AQP1 monomer (28)
and inhibits the water channel when mercury binds (33)
, this observation suggests that CO2 and H2O pass through the same channel. On the other hand, the data in Fig. 3
—showing that DIDS blocks half of the AQP1-dependent, CO2-induced pHS transient in oocytes but has no effect on water permeability—suggests that half of the CO2 passes through AQP1’s central pore and half passes through the four water pores. The absence of an effect of DIDS on water permeability in the present study is consistent with the earlier observation that DIDS does not inhibit the water permeability of red cells (34
, 35)
.
The apparent incongruity of the pCMBS and DIDS data in oocytes could have two explanations that are not mutually exclusive. First, because the earlier intracellular pH measurements (3)
are less sensitive than the current pHS measurements, it is possible that the earlier work could not distinguish a 50% inhibition by pCMBS of CO2 permeability from full blockade. Second, it is possible that the interaction of the large pCMBS molecule with C189 in the mouth of the water pore causes a conformational change that lowers the CO2 permeability of the central pore. We add that, from the molecular dimensions of O2 (effective diameter 2.3–2.9 Å; ref. 30
), it is conceivable that this gas can also pass through the same channel(s), but so far this has not been tested.
Physiological role of the AQP1 pathway for CO2
Recently, Uehlein et al. (7)
have shown that aquaporin of the tobacco plant acts as a CO2 pore and that this pore plays a physiological role in the plant’s uptake of CO2 and growth. The present work is the first demonstration of the physiological role of AQP1 as a channel for a dissolved gas in an intact animal cell. For normal individuals at rest, the O2 and CO2 in pulmonary capillary blood is believed to equilibrate with alveolar air within the first 200 ms of the capillary transit time of 700 ms. To estimate the role of the gas diffusion resistance of the red cell membrane, we have done a simple model calculation in which the intracellular diffusion resistance is estimated by treating the red cell as a plane sheet of 1.6 µm thickness and using the known intracellular CO2 diffusion coefficient (22
, 36)
; to this the inverse of the membrane permeability is added as a second diffusion resistance for CO2. The chemical reaction inside the red cell is assumed to be infinitely fast, which allows one to describe the diffusion and reaction process by replacing the solubility 
in the diffusion equation by ß, a value defined as the change in total intracellular CO2 concentration (CO2 + HCO3–) per CO2 partial pressure difference, i.e., the slope of the CO2 binding curve of red cells (37)
. The time required by Cl–-HCO3– exchange is neglected in this model. We have calculated in this manner the time necessary for CO2 release by a red cell to reach 95% completion (t95%), using the CO2 permeabilities reported here. Because PCO2 values may be somewhat underestimated due to the assumption of perfect intracellular mixing in the calculations of PCO2, the 95% times given in the following may be somewhat overestimated, but this does not severely affect the comparisons made. With a normal PCO2 of 0.15 cm/s, t95% is 110 ms, and is increased to 200 ms when PCO2 assumes a value of 0.06 cm/s in the absence of AQP1. In view of the capillary transit time of 700 ms under resting conditions, it is not expected that AQP1 deficiency causes an impairment of CO2 exchange in the capillary, not even with the reduced transit time that prevails during exercise. However, when the AQP1-independent, DIDS-sensitive pathway for CO2 is also inhibited and PCO2 falls to 0.01 cm/s, one obtains t95% = 990 ms. Thus, if both pathways for molecular CO2 are absent and PCO2 assumes its "basal" value, possibly representing CO2 diffusion across the lipids of the membrane, CO2 release in resting individuals may be less than complete by the end of the pulmonary capillary. Under conditions of exercise, when capillary transit time in the lung is reduced, CO2 release may be subject to a marked diffusion limitation.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
2 Present address: Physiologisches Institut der Universität Zürich, Winterthurerstr. 190, CH-8057 Zürich, Switzerland.
Received for publication March 13, 2006. Accepted for publication May 15, 2006.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  V. Endeward, J.-P. Cartron, P. Ripoche, and G. Gros RhAG protein of the Rhesus complex is a CO2 channel in the human red cell membrane FASEB J, January 1, 2008; 22(1): 64 - 73. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Uehlein and R. Kaldenhoff Aquaporins and Plant Leaf Movements Ann. Bot., January 1, 2008; 101(1): 1 - 4. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Wang and E. Tajkhorshid Molecular Mechanisms of Conduction and Selectivity in Aquaporin Water Channels J. Nutr., June 1, 2007; 137(6): 1509S - 1515S. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. M. Piermarini, E. Y. Kim, and W. F. Boron Evidence against a Direct Interaction between Intracellular Carbonic Anhydrase II and Pure C-terminal Domains of SLC4 Bicarbonate Transporters J. Biol. Chem., January 12, 2007; 282(2): 1409 - 1421. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
5 for each condition; number of donors of Colton null blood was 2. It is seen that PCO2 is significantly reduced when aquaporin-1 is absent. DIDS reduces PCO2 of both normal and Colton null red cells. pCMBS reduces PCO2 of normal but not of Colton null red cells. *Significant differences to the leftmost black column (P<0.05), #Significant differences to the leftmost gray column (P<0.05). Error bars = SD.
