• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of pnasPNASInfo for AuthorsSubscriptionsAboutThis Article
Proc Natl Acad Sci U S A. Mar 19, 2002; 99(6): 3926–3931.
Published online Mar 12, 2002. doi:  10.1073/pnas.062043799
PMCID: PMC122625

Ammonium/methylammonium transport (Amt) proteins facilitate diffusion of NH3 bidirectionally


The ammonium/methylammonium transport (Amt) proteins of enteric bacteria and their homologues, the methylammonium/ammonium permeases of Saccharomyces cerevisiae, are required for fast growth at very low concentrations of the uncharged species NH3. For example, they are essential at low ammonium (NH4+ + NH3) concentrations under acidic conditions. Based on growth studies in batch culture, the Amt protein of Salmonella typhimurium (AmtB) cannot concentrate either NH3 or NH4+ and this organism appears to have no means of doing so. We now show that S. typhimurium releases ammonium into the medium when grown on the alternative nitrogen source arginine and that outward diffusion of ammonium is enhanced by the activity of AmtB. The latter result indicates that AmtB acts bidirectionally. We also confirm a prediction that the AmtB protein would be required at pH 7.0 in ammonium-limited continuous culture, i.e., when the concentration of NH3 is ≤50 nM. Together with our previous studies, current results are in accord with the view that Amt and methylammonium/ammonium permease proteins increase the rate of diffusion of the uncharged species NH3 across the cytoplasmic membrane. These proteins are examples of protein facilitators for a gas.

Members of the ammonium/methylammonium transport (Amt) protein family are found in all three domains of life (1, 2). These cytoplasmic membrane proteins are involved in acquisition of ammonium, often the best nitrogen source for microbes. As a function of pH, ammonium (NH3 + NH4+) exists as a mixture of uncharged (gaseous) and protonated forms. [The pKa of ammonium is 9.25 (3).] Because the uncharged species, NH3, diffuses freely across phospholipid bilayers, there has been controversy about the need for a transport system(s) for ammonium and about the species being transported. Although many reports propose that Amt and methylammonium/ammonium permease (MEP) proteins actively transport the charged species NH4+ (4, 5), there are several lines of evidence against this proposal (6, 7). First, apparent concentration of the ammonium analog, [14C]methylammonium (14CH3NH2 + 14CH3NH3+), in an Amt/MEP-dependent manner, was accounted for by metabolic trapping in enteric bacteria and by diffusion trapping into vacuoles and other acidic compartments in fungi. In each case accumulation depended on a subsequent energy-requiring reaction, one catalyzed by glutamine synthetase in bacteria or the V-type H+-ATPase in fungi. Hence neither example provides evidence for active transport of methylammonium. Second, growth studies in Salmonella typhimurium indicated that the Amt protein of this organism (AmtB) did not concentrate ammonium and that the organism had no means of doing so. Third, in batch culture, amt mutants of enteric bacteria and mep mutants of Saccharomyces cerevisiae had a profound growth defect at low ammonium concentrations only at acid pH, conditions under which the concentration of the uncharged species NH3 is very low. This finding was in accord with the view that NH3 was the substrate for Amt/MEP proteins. Finally, trapping of methylammonium in acidic compartments of fungi is most parsimoniously explained if NH3 is the substrate for Amt/MEP proteins.

We here provide two lines of evidence that the S. typhimurium AmtB protein facilitates diffusion of NH3. First, experiments in batch culture indicate that AmtB enhances leakage of ammonium generated internally when S. typhimurium is grown on the alternative nitrogen source arginine. Hence, the AmtB protein appears to facilitate bidirectional movement of ammonium rather than concentrating it internally. Second, experiments in continuous culture in a chemostat established that AmtB is needed to support normal growth of S. typhimurium under ammonium-limiting conditions even at neutral pH. We previously predicted that continuous culture would be needed to show such a requirement at pH 7 if the substrate for the Amt protein was NH3 rather than NH4+. The evidence to date supports the view that members of the Amt/MEP family facilitate diffusion of a gas. If this is the case, they provide an example of biological gas channels.

Materials and Methods

Media, Growth Conditions, and Strains.

Bacterial strains are listed in Table Table1.1. Cultures were grown at 37°C in appropriately supplemented NC minimal medium essentially as described (6, 8). For experiments at pH 8 or 6, the pH was adjusted with 10 M NaOH or 10 M HCl, respectively. For ammonium-limited chemostats NC medium was supplemented with 27 mM glycerol (0.2%) and 2 mM NH4Cl and for glycerol-limited chemostats it was supplemented with 5 mM glycerol and 10 mM NH4Cl. Delivery of medium from the reservoir to the culture vessel was controlled by a double reciprocating piston pump (HiLoad Pump P-50, Pharmacia). Strain SK3408 (amtB::Camr) (see below) was compared with amtB+ strain SK3539 (putPA1303::[Kans-Camr-′lacZYA]), which carries the same chloramphenicol-resistance gene at an innocuous locus (9). The latter strain does not express the lac genes. Likewise, strain SK3532 (amtB::Camr putPA1303::[Kanr-Φ(glnA′-′lacZYA)]) was compared with SK3041 (putPA1303::[Kanr-Φ(glnA′-′lacZYA)]). Strains carrying the amtB disruption were cultivated in ammonium- and glycerol-limited chemostats five and two times, respectively, whereas amtB+ strains were tested seven and two times, respectively. Chemostat cultures were started with 50 ml of culture freshly grown to full yield in the medium being used. Two samples were taken at each dilution rate by collecting the overflow. The medium in the culture vessel had been replaced at least five times before the first sample was taken and the second sample was taken at ≈24 h. Residual ammonium and glycerol concentrations in the culture vessel, the free pool concentrations of glutamine and glutamate in the cells, and the activity of β-galactosidase in strains carrying glnA-lacZ or amtB-lacZ fusions (Table (Table1)1) were measured as described (8).

Table 1

For crossfeeding experiments NC medium was supplemented with 0.4% glucose and 2.5 mM arginine. Donor and recipient strains were grown on LB medium overnight, and recipient cultures were diluted in unsupplemented NC medium to give 50–200 colonies/plate. Plates were incubated for 2 days to a week at 37°C unless otherwise indicated.

Construction of an amtB::Camr Gene Disruption.

S. typhimurium strain SK3408 was obtained by insertion of a chloramphenicol resistance cassette into the amtB gene of strain SK2979 (wild type) (8), as described below. Oligonucleotides designed for Escherichia coli amtB were used as primers to amplify the entire ORF of the S. typhimurium amtB gene from SK2979. The primers introduced a NheI restriction site at the 5′ end of amtB and a XhoI site at the 3′ end. The 1,281-bp fragment was amplified by PCR with Expand High-Fidelity Taq-polymerase (Roche Molecular Biochemicals) and ligated according to the instructions of the manufacturer into the pNo Ta/T7 shuttle vector (Prime PCR Cloner, 5 Prime→3 Prime) to yield plasmid pJES1149. The fragment was sequenced; at the protein level, S. typhimurium AmtB was 90% identical to E. coli AmtB (data not shown). The 1-kb HincII fragment from plasmid pJES1002, which carries a chloramphenicol resistance cassette, was then ligated into pJES1149, which had been linearized with HpaI, to yield plasmid pJES1151. The 2.3-kb XhoI–NheI fragment from pJES1151, which carries the amtB gene disrupted by the chloramphenicol resistance cassette, was made blunt-ended with the Klenow fragment of DNA polymerase I and moved into the SmaI site of the suicide vector pCVD442 (10) (pJES1034), and the resulting plasmid, pJES1156, was transformed into E. coli DH5α λpir. Plasmid pJES1156, which has the replication origin oriR6K and carries the conditional-lethal gene sacB, was mobilized by tri-parental conjugation into SK3019 (tetracycline resistant) (8) with HB101 carrying pRK2013 (11) (kanamycin resistant) as helper strain. Progeny, which resulted from integration of pJES1156 into the chromosome [it cannot replicate if the host cell does not provide the π protein (12)], were selected on LB plates containing tetracycline (10 μg/ml) and chloramphenicol (25 μg/ml). After overnight growth in LB medium without antibiotics, clones resulting from resolution of the plasmid and integration of the disruption mutation into the amtB gene were selected at 28°C on rich medium (1% tryptone, 0.5% yeast extract, 0.5% NaCl, 1.5% agar) containing 6% sucrose and chloramphenicol. Several clones with the correct phenotype (Amps Tetr Camr Sucr) were further analyzed by PCR amplification and Southern blot analysis (data not shown) to confirm the presence of the amtB::Camr disruption. The amtB::Camr disruption was transferred into other S. typhimurium strains by P22-mediated transduction with selection for Camr.

Construction of a ΔastCADBE::Kanr Deletion/Insertion.

S. typhimurium strain SK3512 (ΔastCADBE::Kanr), which fails to grow on arginine as the nitrogen source, was obtained by deletion of 90% of the ast operon and insertion of a kanamycin resistance cassette (used for selection). The methods were the same as for the construction of the amtB::Camr gene disruption with the following exceptions. Plasmid pJES1303 (see below), which carries the ΔastCADBE::Kanr deletion/insertion, was transformed into the E. coli strain S17–1 λpir and then mobilized by conjugation into SK3408 (chloramphenicol resistant). Progeny were initially selected on LB plates containing chloramphenicol and kanamycin (25 μg/ml). Plasmid pJES1293 was obtained from the shuttle vector pNo Ta/T7 by ligation of a 325-bp PCR fragment that carries the 5 prime end of the astC gene (first gene of the ast operon) (13, 14). Plasmid pJES1294 was obtained from pNo Ta/T7 by ligation of a 311-bp PCR fragment that carries the 3 prime end of the astE gene (last gene of the ast operon). The 1-kb EcoRI fragment from plasmid pUC4K (Pharmacia), which carries a kanamycin resistance cassette, was made blunt-ended and ligated into pJES1293, which had been linearized with EcoRV, to yield pJES1295. The 1.3-kb PmeI fragment from pJES1295 was then ligated into pJES1294, which had been linearized with SmaI, to yield pJES1296. The 1.6-kb SacI–SphI fragment from plasmid pJES1296, which carries the ΔastCADBE::Kanr mutation, was ligated to pCVD442, which had been cleaved with SacI and SphI, to yield pJES1303.


Properties of an amtB Strain of S. typhimurium.

We previously established that the AmtB protein of E. coli was necessary to support rapid growth in batch culture when the concentration of the uncharged species NH3 dropped below 50 nM (6). Like an amtB mutant of E. coli, an S. typhimurium strain carrying a disruption of amtB (SK3408) grew poorly in batch culture at ammonium concentrations ≤1 mM at pH 5 (Table (Table2).2). SK3408 grew normally if the ammonium concentration was increased to 5 mM or the pH was raised to 7. Unlike its parental strain, SK3408 failed to accumulate the ammonium analogue [14C]methylammonium [30 vs. <3 pmol/(ml × OD600 × min) for cells grown on glutamine as nitrogen source (6)]. As reported earlier, expression of amtB was 200-fold higher on glutamine as nitrogen source [differential rate of synthesis of β-galactosidase = 6,500 Miller units for strain SK3362, which carries an amtB-lacZ fusion at the put locus (6)] than on 10 mM NH4Cl.

Table 2
Effect of pH on the growth of an amtB strain of S. typhimurium at different ammonium concentrations

AmtB Enhances Outward Diffusion of Ammonium When S. typhimurium Is Grown on Arginine as the Sole Nitrogen Source.

Although an amtB mutant strain of S. typhimurium had a profound growth defect at low concentrations of NH3 (Table (Table22 and see below), it grew much faster than the congenic wild-type strain on the poor nitrogen source arginine (Table (Table3).3). Thus, AmtB, which is highly expressed on arginine (differential rate of synthesis of β-galactosidase for strain SK3362 = 6,300 Miller units), is deleterious for growth. (It is unlikely that enteric bacteria ever encounter arginine as the sole nitrogen source in a natural environment.) Disruption of amtB also improved growth on arginine in mutant strains with a decreased capacity to assimilate ammonium, e.g., glnA mutant strains in which the catalytic activity of glutamine synthetase is low but not absent (Table (Table3,3, compare strain SK3521 to SK3130) (8). These results led us to hypothesize that AmtB facilitated outward diffusion of ammonium generated intracellularly from arginine.

Table 3
Effect of an amtB disruption and of pH and temperature on the growth of S. typhimurium on arginine

Degradation of arginine by the arginine succinyltransferase (Ast) catabolic pathway, the one present in enteric bacteria, yields 2 mol ammonium and 2 mol glutamate per mol of arginine (13, 14). Our attempts to detect trace amounts of ammonium released into the medium by wild-type (amtB+) strains of S. typhimurium failed and hence we deduce that its concentration did not reach our detection limit of 20 μM. To obtain further evidence that the product being released during growth on arginine was ammonium, we examined the ability of a wild-type strain to crossfeed a mutant strain (SK3512) that was unable to catabolize arginine because it carried a deletion of the entire arginine catabolic (ast) operon. Crossfeeding experiments were performed at pH 7 on plates containing arginine as the sole nitrogen source with a wild-type strain (ast+) (SK2979) as donor and a Δast (SK3512) or Δast amtB strain (SK3514) as recipient. When donor cells were spotted in the middle of the plate, colonies of the Δast mutant closest to the donor grew faster than those farther away (Fig. (Fig.11A). The AmtB protein was required in the recipient because a double mutant strain Δast amtB was crossfed much less well than a single Δast mutant (data not shown). The latter result was in agreement with the view that the nitrogen source provided to the recipient was ammonium. Crossfeeding of the Δast mutant was greatly reduced at 30°C and was reduced at pH 6 (see below). Under both conditions it was completely abolished when the double mutant strain Δast amtB was used as recipient.

Figure 1
Crossfeeding of strains unable to use arginine as a nitrogen source by wild-type and nitrogen regulatory mutant strains. All plates contained arginine as the nitrogen source and glucose as the carbon source and were at pH 7. They were incubated at 37°C. ...

To confirm that ammonium, rather than glutamate, was being excreted when the donor strain was grown on arginine as nitrogen source, we showed that crossfeeding persisted when donor and recipient strains were grown on separate plates. Under these conditions, the substance being crossfed must be able to pass through the gas phase. To perform these tests, we spotted donor cells catabolizing arginine as the sole nitrogen source on only one area of a Petri plate and placed this plate over one that was seeded uniformly with cells of the Δast recipient strain. After a week of incubation at 37°C, colonies of the Δast mutant under the donor were larger than those on the rest of the plate (Fig. (Fig.11B). This process proved that the nitrogen source provided by the donor was ammonium. Crossfeeding was more pronounced if the donor strain carried a nitrogen regulatory mutation [ntrB(Con)] that allows rapid growth on arginine as a nitrogen source (Fig. (Fig.11C). As expected if the substance being provided to the Δast recipient (SK3512) was ammonium, enhanced crossfeeding by the ntrB(Con) donor (SK3004) was eliminated if the Δast recipient also carried a disruption of amtB (SK3514) (Fig. (Fig.11D).

Effects of pH and Temperature Support the View That AmtB Acts Bidirectionally.

Growth of a wild-type strain (SK2979) on arginine as the sole nitrogen source strongly depended on the pH of the medium. Growth rate decreased with pH, from a relatively fast growth rate at pH 8 to a much slower one at pH 6 (Table (Table3).3). This effect was less pronounced for the amtB mutant strain SK3408 and was specific to arginine because it did not occur with ammonium as the sole nitrogen source (Table (Table44 and strain SK2979 in Table Table2).2). We postulate that decreases in growth rate on arginine at low pH may be caused by acidic trapping of NH3 in the medium. Such trapping would enhance outward diffusion of NH3 by keeping its external concentration lower than its internal concentration. (For a converse situation in fungi see ref. 7.)

Table 4
Effect of an amtB disruption and of pH and temperature on the growth of S. typhimurium on ammonium

In contrast to the case for ammonium (Table (Table4),4), growth of the wild-type strain on arginine as the nitrogen source was faster at low than high temperature (Table (Table3).3). Growth of the amtB mutant strain was essentially unaffected, demonstrating that temperature effects depended on AmtB rather than arginine catabolic enzymes. Lower temperature would slow outward diffusion of NH3 and would increase its solubility in the medium. One or both of these factors might limit loss of NH3 from cells at low temperature and thereby contribute to faster growth.

An amtB Strain of S. typhimurium Has a Growth Defect in Ammonium-Limited Continuous Culture at pH 7.0.

To yield ≤50 nM NH3, the concentration at which AmtB is required for rapid growth (Table (Table2;2; ref. 6), the total ammonium concentration must be decreased to ≤10 μM at pH 7. This is too little total nitrogen to give a detectable yield in batch culture and hence we tested growth of the amtB mutant of S. typhimurium and amtB+ strains in ammonium-limited continuous culture (Materials and Methods). In different experiments amtB+ strains were stable at dilution rates D ≤0.59 h−1 to 0.84 h−1 (see also ref. 8). Dilution rates of 0.59 and 0.84 h−1 correspond to doubling times of 70 and 50 min, respectively, and we observed doubling times over this range in different experiments in batch culture [minimal medium with glycerol (27 or 5 mM) as carbon source and ammonium (2 or 10 mM) as nitrogen source]. In an early experiment (Fig. (Fig.22A), in which we increased the dilution rate from 0.34 h−1 to rates near maximal, the cell yield of an amtB+ strain carrying a glnA-lacZ fusion at the put locus (SK3041) was constant (OD600 of 0.57) until the culture washed out. As washout was approached the glutamine pool concentration increased more than 15-fold with a characteristic spike (Fig. (Fig.22B; see also Fig. Fig.22D and ref. 8). As expected (8), expression of glnA-lacZ decreased. By contrast, the cell yield of an amtB disruption strain carrying the glnA-lacZ fusion (SK3532) steadily decreased as D was increased toward maximal, and the culture began to foam. Although the glutamine pool concentration rose at washout (see also Fig. Fig.22D), the spike characteristic of amtB+ strains was not observed, nor was there a decrease in glnA-lacZ expression. The latter occurred abruptly at washout. Presumably, differences for the amtB strain reflect its difficulty in obtaining ammonium. Foaming of strains carrying the amtB disruption was a problem in five independent ammonium-limited chemostats, two of which are described below.

Figure 2
Strains with an amtB lesion have a growth defect in ammonium-limited but not glycerol-limited chemostats. The amtB+ and amtB mutant strains for A and B were SK3041 (glnA-lacZ) and SK3532 (amtB::Camr glnA-lacZ), respectively, and those for ...

To determine whether the cell density of an amtB strain could be stabilized in an ammonium-limited chemostat, we started and maintained strain SK3408 (amtB::Camr) at a slow dilution rate (D = 0.34 h−1, which corresponds to a doubling time of 120 min). Over a period of 4 days, the culture washed out and foamed (Fig. (Fig.22C). The glutamine pool concentration increased and the glutamate pool concentration rose to abnormally high values (Fig. (Fig.22D). The culture appeared to stabilize when D was decreased to 0.09 h−1 (calculated doubling time of 7.7 h). However, this condition is essentially equivalent to a batch culture in stationary phase. The same behavior was observed when strain SK3408 was started at D = 0.34 h−1 and then shifted up to D = 0.41, 0.45, and 0.49 h−1 (not shown). After 5 days total, the culture foamed and washed out (at D = 0.49 h−1; calculated doubling time of 84 min).

To confirm that the peculiar behavior of the amtB strain was unique to it, we maintained an amtB+ strain carrying a chloramphenicol resistance cassette at the put locus (SK3539) in an ammonium-limited chemostat for 23 days (Fig. (Fig.22C). The chemostat was started at D = 0.34 h−1 and D was then shifted up from and back to this value over the course of the experiment. The responses of the glutamine and glutamate pools (Fig. (Fig.22D) were essentially the same as those described for the experiment of Fig. Fig.22 A and B and experiments reported previously (8). (Note that the spike in the glutamine pool concentration is compressed in Fig. Fig.22D because of the change in the x axis.) For both the amtB+ strain and the amtB mutant (e.g., in the experiments of Fig. Fig.22 AD), residual ammonium could be detected in the culture vessel only during washout and the glycerol concentration rose at this point (not shown). Our limit of detection for both was 20 μM.

The amtB Strain Has No Growth Defect in Glycerol-Limited Continuous Culture.

Because ammonium appears to be acquired in the same way as glycerol—that is, unmediated diffusion at high concentrations and facilitated diffusion at low concentrations—we compared the behavior of amtB and wild-type strains in glycerol-limited continuous culture at pH 7 as a control. Under these conditions the amtB lesion had no effect. As observed previously (8, 15), the cell yield (OD600) for both strains increased slightly with increasing dilution rate up to D of 0.59 h−1 (Fig. (Fig.22E). At D <0.59 h−1, the concentration of glycerol that remained in the culture medium of either strain was below our detection limit of 20 μM, whereas the concentration of ammonium remained high (≥7 mM) (not shown). Glutamate pool concentrations were steady and high, whereas glutamine pools were lower (Fig. (Fig.22F). Both cultures washed out at D values >0.59 h−1, which correspond to doubling times faster than 70 min. Under these conditions, the residual glycerol and ammonium concentrations remaining in the medium increased.


AmtB Facilitates Bidirectional Movement of Ammonium.

The failure of an amtB mutant of S. typhimurium to stabilize in ammonium-limited continuous culture at pH 7 (Fig. (Fig.22 AD) indicates that the requirement for AmtB is not restricted to acid pH (Table (Table2).2). Rather the function of this protein is needed whenever the external concentration of NH3 declines to the 50-nM range. At concentrations ≤50 nM the rate of unmediated diffusion of NH3 across the cytoplasmic membrane of enteric bacteria is apparently not fast enough for optimal growth. Together with previous results (6, 7), the present findings are readily explained if the substrate for AmtB is NH3 rather than NH4+. In contrast to the case for strains with a defect in the high-affinity pathway for assimilation of ammonium (gltB or ΔgltB) (ref. 6 and D. Yan and S.K., unpublished work), strains carrying an amtB disruption apparently cannot acquire suppressor mutations that stabilize their growth in an ammonium-limited chemostat. The failure of S. typhimurium to compensate for the loss of AmtB gives a hint that some aspect of AmtB function is unique (16) (see below).

Improvement of growth on the alternative nitrogen source arginine in strains carrying an amtB disruption (Table (Table3)3) provides evidence that AmtB can increase the rate of outward diffusion of ammonium that is generated internally. That it is indeed ammonium that is leaked when arginine is used as the sole nitrogen source was verified by the ability of wild-type and nitrogen regulatory mutant strains [ntrB(Con)] to crossfeed strains that lack the entire arginine catabolic operon (Δast), even when donor and recipient were on separate Petri plates (Fig. (Fig.11 BD). Under the latter circumstances the substance being crossfed must pass through the gas phase. As expected, crossfeeding of the Δast recipient strain was less effective if it also carried an amtB disruption. Effects of an amtB disruption on utilization of ammonium generated internally as well as on use of external ammonium indicate that Amt proteins facilitate bidirectional movement of ammonium. Thus, they are ammonium channels rather than active transporters.

Quantitative Role of AmtB.

At 25 nM external NH3—a condition achieved in batch culture at 0.5 mM ammonium and pH 5—disruption of amtB results in about a 4-fold decrease in growth rate (4-fold increase in doubling time) (Table (Table2).2). The growth rate of the wild-type strain is just beginning to decrease under these conditions, indicating that it no longer has excess capacity to acquire ammonium. Together, these observations provide evidence that a large proportion (we estimate 3/4) of the ammonium entering cells at 25 nM NH3 is entering through AmtB. (Uncertainty is caused by the fact that the ammonium concentration and the growth rate are constantly decreasing during the course of the experiment.) The remainder of the ammonium is presumably crossing the cytoplasmic membrane in an unmediated manner and this portion must be crossing as NH3. In our ammonium-limited chemostat cultures the actual concentration of ammonium may have been as low as 1 μM [calculated on the assumption that each drop of fresh medium (= 0.2 ml) containing 2 mM ammonium was uniformly mixed in the 320 ml volume of the chemostat]. At pH 7 this corresponds to an NH3 concentration of 5 nM. Thus, AmtB appears to function even when the concentration of NH3 declines to the nM range. The classic kinetic studies of Hackette and colleagues (17) on ammonium transport in Penicillium chrysogenum lend credence to this view. Hackette and colleagues noted that if the substrate of the transporter was NH3 rather than NH4+, the Km of the Penicillium transporter would have to be ≈0.3 nM.

Biological Gas Channels.

If indeed, the substrate for the bidirectional facilitators Amt/MEP is NH3, these proteins are biological gas channels. Although the existence of gas channels has been postulated (e.g., ref. 18 and 19), to our knowledge, Amt/MEP proteins provide the first example. Despite their widespread distribution among organisms and presumably ancient origin (1, 2, 16), Amt/MEP proteins show homology only to Rhesus (Rh) proteins (2), of which the human Rh blood group substance is best known. We have previously speculated that Rh and Amt/MEP proteins have in common that they are facilitators for gas transport (7). However, we postulate that the substrate for Rh proteins is CO2, a hypothesis that is commensurate with the relatively restricted organismal, organ, and tissue distribution of Rh proteins (20) and with the fact that the abundance of the Rh blood group substance in red cell membranes is second only to that of the bicarbonate (anion) exchanger (band 3).


We thank Laszlo Csonka, Boris Magasanik, and Hiroshi Nikaido for critical review of the manuscript. This work was supported by National Science Foundation Grant MCB 9874443 to S.K.


ammonium/methylammonium transport
methylammonium/ammonium permease
arginine succinyltransferase


1. Howitt S M, Udvardi M K. Biochim Biophys Acta. 2000;1465:152–170. [PubMed]
2. Marini A-M, Urrestarazu A, Beauwens R, Bruno A. Trends Biochem Sci. 1997;22:460–461. [PubMed]
3. Lide D R. Handbook of Chemistry and Physics. Cleveland: Chemical Rubber Publishing; 2001.
4. Marini A-M, Soussi-Boudekou S, Vissers S, André B. Mol Cell Biol. 1997;17:4282–4293. [PMC free article] [PubMed]
5. von Wiren N, Gazzarrini S, Gojon A, Frommer W B. Curr Opin Plant Biol. 2000;3:254–261. [PubMed]
6. Soupene E, He L, Yan D, Kustu S. Proc Natl Acad Sci USA. 1998;95:7030–7034. [PMC free article] [PubMed]
7. Soupene E, Ramirez R M, Kustu S. Mol Cell Biol. 2001;21:5733–5741. [PMC free article] [PubMed]
8. Ikeda T P, Shauger A E, Kustu S. J Mol Biol. 1996;259:589–607. [PubMed]
9. Elliott T. J Bacteriol. 1992;174:245–253. [PMC free article] [PubMed]
10. Donnenberg M S, Kaper J B. Infect Immun. 1991;59:4310–4317. [PMC free article] [PubMed]
11. Ditta G, Stanfield S, Corbin D, Helinski D R. Proc Natl Acad Sci USA. 1980;77:7347–7351. [PMC free article] [PubMed]
12. Kolter R, Inuzuka M, Helinski D R. Cell. 1978;15:1199–1208. [PubMed]
13. Itoh Y. J Bacteriol. 1997;179:7280–7290. [PMC free article] [PubMed]
14. Schneider B L, Kiupakis A K, Reitzer L J. J Bacteriol. 1998;180:4278–4286. [PMC free article] [PubMed]
15. Kubitschek H E. Introduction to Research with Continuous Cultures. Englewood Cliffs, NJ: Prentice–Hall; 1970.
16. Pao S S, Paulsen I T, Saier M H., Jr Microbiol Mol Biol Rev. 1998;62:1–34. [PMC free article] [PubMed]
17. Hackette S L, Skye G E, Burton C, Segel I H. J Biol Chem. 1970;245:4241–4250. [PubMed]
18. Forster R E, Gros G, Lin L, Ono Y, Wunder M. Proc Natl Acad Sci USA. 1998;95:15815–15820. [PMC free article] [PubMed]
19. Pawloski J R, Hess D T, Stamler J S. Nature (London) 2001;409:622–626. [PubMed]
20. Huang C H, Liu P Z. Blood Cells Mol Dis. 2001;27:90–101. [PubMed]

Articles from Proceedings of the National Academy of Sciences of the United States of America are provided here courtesy of National Academy of Sciences
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...